# vidbenchmark - full AI video model catalog

> Machine-readable catalog of every AI video model tracked by vidbenchmark.com. Blind same-prompt Elo scores, eight-axis ratings and per-provider pricing. Catalog updated 2026-08-04; benchmarks re-run weekly.

Scores are 0-100 and come from blind, vendor-anonymised human ratings on an identical prompt set. Elo comes from head-to-head matchups between those renders. Prices are list prices per second of finished output, collected per provider.

## 1. MiniMax H3

- Page: https://vidbenchmark.com/model/minimax-h3
- Vendor: MiniMax
- Released: 2026-05-28
- Elo: 1268 (-4 over 7 days)
- Price: $0.13/second
- Max resolution: 2K
- Max duration: 15s
- Native audio: yes
- Lipsync: yes
- Character reference: yes
- Open weights: no
- Deprecated: no
- Best for: Cinema quality + audio

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 96 |
| Motion realism | 93 |
| Character consistency | 95 |
| Temporal coherence | 96 |
| Lipsync accuracy | 89 |
| Native audio quality | 96 |
| Speed to first frame | 53 |
| Cost efficiency | 48 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.078 | 56s |
| Novita | $0.114 | 65s |
| Replicate | $0.146 | 70s |

### Sample prompts

- Hey, this is a video sample from vidbenchmark dot com testing AI models side by side.

## 2. Seedance 2.0

- Page: https://vidbenchmark.com/model/seedance-2
- Vendor: ByteDance
- Released: 2026-06-15
- Elo: 1261 (+29 over 7 days)
- Price: $0.151/second
- Max resolution: 1080p
- Max duration: 12s
- Native audio: yes
- Lipsync: yes
- Character reference: yes
- Open weights: no
- Deprecated: no
- Best for: Character-driven narrative

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 100 |
| Motion realism | 96 |
| Character consistency | 95 |
| Temporal coherence | 96 |
| Lipsync accuracy | 93 |
| Native audio quality | 91 |
| Speed to first frame | 67 |
| Cost efficiency | 46 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.091 | 50s |
| Novita | $0.133 | 58s |

### Sample prompts

- Hey, this is a video sample from vidbenchmark dot com testing AI models side by side.

## 3. Gemini Omni Flash

- Page: https://vidbenchmark.com/model/gemini-omni-flash
- Vendor: Google
- Released: 2026-07-09
- Elo: 1254 (+9 over 7 days)
- Price: $0.1/second
- Max resolution: 1080p
- Max duration: 8s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: no
- Deprecated: no
- Best for: Fast prompt adherence

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 100 |
| Motion realism | 94 |
| Character consistency | 92 |
| Temporal coherence | 94 |
| Lipsync accuracy | 93 |
| Native audio quality | 89 |
| Speed to first frame | 84 |
| Cost efficiency | 54 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.06 | 42s |
| Novita | $0.088 | 48s |
| fal.ai | $0.1 | 36s |
| Replicate | $0.112 | 52s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Gemini Omni Flash

## 4. Grok Imagine 1.5

- Page: https://vidbenchmark.com/model/grok-imagine-1-5
- Vendor: xAI
- Released: 2026-06-02
- Elo: 1241 (+29 over 7 days)
- Price: $0.14/second
- Max resolution: 1080p
- Max duration: 6s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: no
- Deprecated: no
- Best for: Creative freedom

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 91 |
| Motion realism | 89 |
| Character consistency | 92 |
| Temporal coherence | 94 |
| Lipsync accuracy | 84 |
| Native audio quality | 85 |
| Speed to first frame | 72 |
| Cost efficiency | 45 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| fal.ai | $0.14 | 33s |
| RunningHub | $0.143 | 43s |
| Segmind | $0.175 | 37s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Grok Imagine 1.5

## 5. HappyHorse 1.1

- Page: https://vidbenchmark.com/model/happyhorse-1-1
- Vendor: Alibaba
- Released: 2026-05-11
- Elo: 1226 (-14 over 7 days)
- Price: $0.165/second
- Max resolution: 1080p
- Max duration: 8s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: no
- Deprecated: no
- Best for: Natural movement

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 92 |
| Motion realism | 93 |
| Character consistency | 82 |
| Temporal coherence | 90 |
| Lipsync accuracy | 84 |
| Native audio quality | 91 |
| Speed to first frame | 60 |
| Cost efficiency | 47 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Novita | $0.145 | 51s |
| Replicate | $0.185 | 56s |
| Segmind | $0.206 | 42s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on HappyHorse 1.1

## 6. Wan 2.7

- Page: https://vidbenchmark.com/model/wan-2-7
- Vendor: Alibaba
- Released: 2026-06-20
- Elo: 1215 (-11 over 7 days)
- Price: $0.15/second
- Max resolution: 1080p
- Max duration: 10s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: yes
- Deprecated: no
- Best for: Best open-weight w/ audio

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 93 |
| Motion realism | 86 |
| Character consistency | 84 |
| Temporal coherence | 87 |
| Lipsync accuracy | 86 |
| Native audio quality | 88 |
| Speed to first frame | 64 |
| Cost efficiency | 48 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.09 | 50s |
| Novita | $0.132 | 57s |
| fal.ai | $0.15 | 42s |

### Sample prompts

- Hey, this is a video sample from vidbenchmark dot com testing AI models side by side.

## 7. SkyReels V4

- Page: https://vidbenchmark.com/model/skyreels-v4
- Vendor: Skywork
- Released: 2026-04-30
- Elo: 1206 (-10 over 7 days)
- Price: $0.35/second
- Max resolution: 1080p
- Max duration: 10s
- Native audio: yes
- Lipsync: yes
- Character reference: yes
- Open weights: no
- Deprecated: no
- Best for: Long-form storytelling

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 87 |
| Motion realism | 87 |
| Character consistency | 91 |
| Temporal coherence | 90 |
| Lipsync accuracy | 84 |
| Native audio quality | 85 |
| Speed to first frame | 67 |
| Cost efficiency | 31 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Novita | $0.308 | 58s |
| Replicate | $0.392 | 63s |
| Segmind | $0.438 | 48s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on SkyReels V4

## 8. Veo 3.1

- Page: https://vidbenchmark.com/model/veo-3-1
- Vendor: Google
- Released: 2026-03-18
- Elo: 1198 (+6 over 7 days)
- Price: $0.4/second
- Max resolution: 1080p
- Max duration: 8s
- Native audio: yes
- Lipsync: yes
- Character reference: yes
- Open weights: no
- Deprecated: no
- Best for: Prompt adherence king

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 92 |
| Motion realism | 82 |
| Character consistency | 87 |
| Temporal coherence | 87 |
| Lipsync accuracy | 85 |
| Native audio quality | 82 |
| Speed to first frame | 71 |
| Cost efficiency | 35 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Novita | $0.352 | 54s |
| Replicate | $0.448 | 59s |
| Segmind | $0.5 | 45s |

### Sample prompts

- Hey, this is a video sample from vidbenchmark dot com testing AI models side by side.

## 9. Grok Imagine

- Page: https://vidbenchmark.com/model/grok-imagine
- Vendor: xAI
- Released: 2025-11-14
- Elo: 1188 (+16 over 7 days)
- Price: $0.07/second
- Max resolution: 1080p
- Max duration: 6s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: no
- Deprecated: no
- Best for: Grok v1

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 85 |
| Motion realism | 88 |
| Character consistency | 77 |
| Temporal coherence | 84 |
| Lipsync accuracy | 83 |
| Native audio quality | 79 |
| Speed to first frame | 65 |
| Cost efficiency | 58 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.042 | 44s |
| Novita | $0.062 | 50s |
| Replicate | $0.078 | 54s |
| Segmind | $0.088 | 41s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Grok Imagine

## 10. Veo 3.1 Fast

- Page: https://vidbenchmark.com/model/veo-3-1-fast
- Vendor: Google
- Released: 2026-03-18
- Elo: 1181 (+20 over 7 days)
- Price: $0.15/second
- Max resolution: 720p
- Max duration: 8s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: no
- Deprecated: no
- Best for: Cheap Google option

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 83 |
| Motion realism | 84 |
| Character consistency | 84 |
| Temporal coherence | 82 |
| Lipsync accuracy | 80 |
| Native audio quality | 78 |
| Speed to first frame | 86 |
| Cost efficiency | 48 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| RunningHub | $0.153 | 53s |
| Replicate | $0.168 | 61s |
| Segmind | $0.188 | 46s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Veo 3.1 Fast

## 11. Kling 3.0 Pro 1080p

- Page: https://vidbenchmark.com/model/kling-3-0-pro-1080p
- Vendor: KlingAI
- Released: 2026-04-08
- Elo: 1173 (+19 over 7 days)
- Price: $0.336/second
- Max resolution: 1080p
- Max duration: 10s
- Native audio: yes
- Lipsync: yes
- Character reference: yes
- Open weights: no
- Deprecated: no
- Best for: Professional Kling tier

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 84 |
| Motion realism | 84 |
| Character consistency | 89 |
| Temporal coherence | 80 |
| Lipsync accuracy | 78 |
| Native audio quality | 76 |
| Speed to first frame | 66 |
| Cost efficiency | 32 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| fal.ai | $0.336 | 46s |
| RunningHub | $0.343 | 59s |
| Replicate | $0.376 | 67s |
| Segmind | $0.42 | 51s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Kling 3.0 Pro 1080p

## 12. PixVerse V6

- Page: https://vidbenchmark.com/model/pixverse-v6
- Vendor: PixVerse
- Released: 2026-06-25
- Elo: 1165 (+5 over 7 days)
- Price: $0.115/second
- Max resolution: 1080p
- Max duration: 15s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: no
- Deprecated: no
- Best for: Anime + stylized

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 84 |
| Motion realism | 86 |
| Character consistency | 75 |
| Temporal coherence | 83 |
| Lipsync accuracy | 76 |
| Native audio quality | 83 |
| Speed to first frame | 66 |
| Cost efficiency | 48 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.069 | 65s |
| Novita | $0.101 | 75s |
| fal.ai | $0.115 | 56s |
| Replicate | $0.129 | 82s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on PixVerse V6

## 13. Kling 3.0 Standard 720p

- Page: https://vidbenchmark.com/model/kling-3-0-standard-720p
- Vendor: KlingAI
- Released: 2026-04-08
- Elo: 1156 (+9 over 7 days)
- Price: $0.26/second
- Max resolution: 720p
- Max duration: 10s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: no
- Deprecated: no
- Best for: Kling economy

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 79 |
| Motion realism | 82 |
| Character consistency | 75 |
| Temporal coherence | 82 |
| Lipsync accuracy | 72 |
| Native audio quality | 78 |
| Speed to first frame | 65 |
| Cost efficiency | 37 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.156 | 55s |
| fal.ai | $0.26 | 47s |
| RunningHub | $0.265 | 61s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Kling 3.0 Standard 720p

## 14. Kling 3 Omni 1080p

- Page: https://vidbenchmark.com/model/kling-3-omni-1080p
- Vendor: KlingAI
- Released: 2026-07-02
- Elo: 1148 (+28 over 7 days)
- Price: $0.28/second
- Max resolution: 1080p
- Max duration: 15s
- Native audio: yes
- Lipsync: yes
- Character reference: yes
- Open weights: no
- Deprecated: no
- Best for: Voice-driven multi-scene

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 82 |
| Motion realism | 75 |
| Character consistency | 81 |
| Temporal coherence | 81 |
| Lipsync accuracy | 71 |
| Native audio quality | 72 |
| Speed to first frame | 68 |
| Cost efficiency | 39 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.168 | 67s |
| fal.ai | $0.28 | 57s |
| RunningHub | $0.286 | 74s |
| Segmind | $0.35 | 64s |

### Sample prompts

- Hey, this is a video sample from vidbenchmark dot com testing AI models side by side.

## 15. Vidu Q3 Pro

- Page: https://vidbenchmark.com/model/vidu-q3-pro
- Vendor: Vidu
- Released: 2026-05-19
- Elo: 1141 (-8 over 7 days)
- Price: $0.16/second
- Max resolution: 1080p
- Max duration: 16s
- Native audio: yes
- Lipsync: yes
- Character reference: yes
- Open weights: no
- Deprecated: no
- Best for: 16s length record

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 84 |
| Motion realism | 75 |
| Character consistency | 78 |
| Temporal coherence | 74 |
| Lipsync accuracy | 72 |
| Native audio quality | 73 |
| Speed to first frame | 72 |
| Cost efficiency | 46 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.096 | 70s |
| Novita | $0.141 | 80s |
| Replicate | $0.179 | 87s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Vidu Q3 Pro

## 16. Kling 3 Omni 720p Std

- Page: https://vidbenchmark.com/model/kling-3-omni-720p-std
- Vendor: KlingAI
- Released: 2026-07-02
- Elo: 1130 (+19 over 7 days)
- Price: $0.224/second
- Max resolution: 720p
- Max duration: 15s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: no
- Deprecated: no
- Best for: Kling voice-driven cheap

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 78 |
| Motion realism | 79 |
| Character consistency | 77 |
| Temporal coherence | 76 |
| Lipsync accuracy | 70 |
| Native audio quality | 77 |
| Speed to first frame | 64 |
| Cost efficiency | 41 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| fal.ai | $0.224 | 58s |
| RunningHub | $0.228 | 76s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Kling 3 Omni 720p Std

## 17. Runway Gen-4 Turbo

- Page: https://vidbenchmark.com/model/runway-gen-4
- Vendor: Runway
- Released: 2026-02-27
- Elo: 1122 (+16 over 7 days)
- Price: $0.18/second
- Max resolution: 1080p
- Max duration: 10s
- Native audio: no
- Lipsync: no
- Character reference: yes
- Open weights: no
- Deprecated: no
- Best for: Character consistency + editing

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 81 |
| Motion realism | 80 |
| Character consistency | 77 |
| Temporal coherence | 76 |
| Lipsync accuracy | 0 |
| Native audio quality | 0 |
| Speed to first frame | 67 |
| Cost efficiency | 45 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| fal.ai | $0.18 | 50s |
| RunningHub | $0.184 | 64s |
| Replicate | $0.202 | 73s |
| Segmind | $0.225 | 55s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Runway Gen-4 Turbo

## 18. Sora 2 Pro

- Page: https://vidbenchmark.com/model/sora-2-pro
- Vendor: OpenAI
- Released: 2025-12-09
- Elo: 1113 (+23 over 7 days)
- Price: $0.5/second
- Max resolution: 1080p
- Max duration: 20s
- Native audio: yes
- Lipsync: yes
- Character reference: yes
- Open weights: no
- Deprecated: yes
- Best for: Deprecated Sep 2026

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 76 |
| Motion realism | 79 |
| Character consistency | 81 |
| Temporal coherence | 70 |
| Lipsync accuracy | 73 |
| Native audio quality | 75 |
| Speed to first frame | 71 |
| Cost efficiency | 29 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Novita | $0.44 | 93s |
| Replicate | $0.56 | 101s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Sora 2 Pro

## 19. Sora 2

- Page: https://vidbenchmark.com/model/sora-2
- Vendor: OpenAI
- Released: 2025-12-09
- Elo: 1098 (-4 over 7 days)
- Price: $0.1/second
- Max resolution: 720p
- Max duration: 10s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: no
- Deprecated: yes
- Best for: Deprecated Sep 2026

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 76 |
| Motion realism | 71 |
| Character consistency | 68 |
| Temporal coherence | 75 |
| Lipsync accuracy | 64 |
| Native audio quality | 66 |
| Speed to first frame | 66 |
| Cost efficiency | 51 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.06 | 61s |
| fal.ai | $0.1 | 51s |
| RunningHub | $0.102 | 67s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Sora 2

## 20. Kling 2.6 Pro

- Page: https://vidbenchmark.com/model/kling-2-6-pro
- Vendor: KlingAI
- Released: 2025-10-21
- Elo: 1090 (-16 over 7 days)
- Price: $0.14/second
- Max resolution: 1080p
- Max duration: 10s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: no
- Deprecated: no
- Best for: Prior-gen Kling

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 76 |
| Motion realism | 72 |
| Character consistency | 71 |
| Temporal coherence | 71 |
| Lipsync accuracy | 69 |
| Native audio quality | 66 |
| Speed to first frame | 64 |
| Cost efficiency | 47 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| fal.ai | $0.14 | 52s |
| RunningHub | $0.143 | 67s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Kling 2.6 Pro

## 21. Seedance 1.5 Pro

- Page: https://vidbenchmark.com/model/seedance-1-5-pro
- Vendor: ByteDance
- Released: 2025-12-02
- Elo: 1081 (-14 over 7 days)
- Price: $0.198/second
- Max resolution: 1080p
- Max duration: 10s
- Native audio: no
- Lipsync: no
- Character reference: yes
- Open weights: no
- Deprecated: no
- Best for: Prior-gen Seedance

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 72 |
| Motion realism | 72 |
| Character consistency | 72 |
| Temporal coherence | 70 |
| Lipsync accuracy | 0 |
| Native audio quality | 0 |
| Speed to first frame | 70 |
| Cost efficiency | 39 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.119 | 62s |
| Novita | $0.174 | 71s |
| fal.ai | $0.198 | 53s |
| Replicate | $0.222 | 78s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Seedance 1.5 Pro

## 22. Cosmos3-Super-Image2Video

- Page: https://vidbenchmark.com/model/cosmos3
- Vendor: NVIDIA
- Released: 2026-05-06
- Elo: 1070 (+5 over 7 days)
- Price: $0.05/second
- Max resolution: 1080p
- Max duration: 6s
- Native audio: no
- Lipsync: no
- Character reference: no
- Open weights: yes
- Deprecated: no
- Best for: Open + strong image2video

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 75 |
| Motion realism | 67 |
| Character consistency | 66 |
| Temporal coherence | 69 |
| Lipsync accuracy | 0 |
| Native audio quality | 0 |
| Speed to first frame | 60 |
| Cost efficiency | 66 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| RunningHub | $0.051 | 60s |
| Replicate | $0.056 | 68s |
| Segmind | $0.063 | 52s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Cosmos3-Super-Image2Video

## 23. LTX-2.3 Pro

- Page: https://vidbenchmark.com/model/ltx-2-3-pro
- Vendor: Lightricks
- Released: 2026-06-11
- Elo: 1058 (+1 over 7 days)
- Price: $0.08/second
- Max resolution: 1080p
- Max duration: 6s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: yes
- Deprecated: no
- Best for: Open-weight quality

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 68 |
| Motion realism | 65 |
| Character consistency | 59 |
| Temporal coherence | 62 |
| Lipsync accuracy | 60 |
| Native audio quality | 67 |
| Speed to first frame | 60 |
| Cost efficiency | 56 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.048 | 55s |
| Novita | $0.07 | 64s |
| fal.ai | $0.08 | 47s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on LTX-2.3 Pro

## 24. LTX-2.3 Fast

- Page: https://vidbenchmark.com/model/ltx-2-3-fast
- Vendor: Lightricks
- Released: 2026-06-11
- Elo: 1047 (-20 over 7 days)
- Price: $0.04/second
- Max resolution: 1080p
- Max duration: 6s
- Native audio: yes
- Lipsync: no
- Character reference: no
- Open weights: yes
- Deprecated: no
- Best for: Cheapest usable

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 71 |
| Motion realism | 66 |
| Character consistency | 63 |
| Temporal coherence | 61 |
| Lipsync accuracy | 59 |
| Native audio quality | 62 |
| Speed to first frame | 88 |
| Cost efficiency | 66 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| fal.ai | $0.04 | 48s |
| RunningHub | $0.041 | 62s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on LTX-2.3 Fast

## 25. PixVerse V5.6

- Page: https://vidbenchmark.com/model/pixverse-v5-6
- Vendor: PixVerse
- Released: 2025-12-16
- Elo: 1036 (+16 over 7 days)
- Price: $0.115/second
- Max resolution: 1080p
- Max duration: 8s
- Native audio: yes
- Lipsync: yes
- Character reference: no
- Open weights: no
- Deprecated: no
- Best for: PixVerse prior-gen

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 63 |
| Motion realism | 65 |
| Character consistency | 60 |
| Temporal coherence | 59 |
| Lipsync accuracy | 61 |
| Native audio quality | 63 |
| Speed to first frame | 66 |
| Cost efficiency | 50 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Novita | $0.101 | 71s |
| RunningHub | $0.117 | 68s |
| Replicate | $0.129 | 77s |
| Segmind | $0.144 | 59s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on PixVerse V5.6

## 26. Hailuo 2.3 Pro

- Page: https://vidbenchmark.com/model/hailuo-2-3-pro
- Vendor: MiniMax
- Released: 2025-09-30
- Elo: 1025 (-28 over 7 days)
- Price: $0.1/second
- Max resolution: 1080p
- Max duration: 8s
- Native audio: no
- Lipsync: no
- Character reference: no
- Open weights: no
- Deprecated: no
- Best for: MiniMax classic

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 68 |
| Motion realism | 65 |
| Character consistency | 59 |
| Temporal coherence | 66 |
| Lipsync accuracy | 0 |
| Native audio quality | 0 |
| Speed to first frame | 68 |
| Cost efficiency | 50 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| fal.ai | $0.1 | 53s |
| RunningHub | $0.102 | 69s |
| Segmind | $0.125 | 60s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Hailuo 2.3 Pro

## 27. LTX-2 Fast

- Page: https://vidbenchmark.com/model/ltx-2-fast
- Vendor: Lightricks
- Released: 2025-11-05
- Elo: 1010 (-26 over 7 days)
- Price: $0.04/second
- Max resolution: 720p
- Max duration: 6s
- Native audio: no
- Lipsync: no
- Character reference: no
- Open weights: yes
- Deprecated: no
- Best for: Speed-optimized open

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 61 |
| Motion realism | 61 |
| Character consistency | 53 |
| Temporal coherence | 60 |
| Lipsync accuracy | 0 |
| Native audio quality | 0 |
| Speed to first frame | 93 |
| Cost efficiency | 66 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| RunningHub | $0.041 | 66s |
| Replicate | $0.045 | 75s |
| Segmind | $0.05 | 57s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on LTX-2 Fast

## 28. Agnes-Video-V2

- Page: https://vidbenchmark.com/model/agnes-video-v2
- Vendor: Sapiens AI
- Released: 2026-01-22
- Elo: 995 (+4 over 7 days)
- Price: $0.005/second
- Max resolution: 720p
- Max duration: 6s
- Native audio: no
- Lipsync: no
- Character reference: no
- Open weights: no
- Deprecated: no
- Best for: Rock-bottom price

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 59 |
| Motion realism | 61 |
| Character consistency | 59 |
| Temporal coherence | 58 |
| Lipsync accuracy | 0 |
| Native audio quality | 0 |
| Speed to first frame | 72 |
| Cost efficiency | 99 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.003 | 61s |
| fal.ai | $0.005 | 52s |
| RunningHub | $0.005 | 67s |
| Segmind | $0.006 | 58s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Agnes-Video-V2

## 29. Wan 2.6

- Page: https://vidbenchmark.com/model/wan-2-6
- Vendor: Alibaba
- Released: 2025-10-08
- Elo: 978 (+1 over 7 days)
- Price: $0.15/second
- Max resolution: 1080p
- Max duration: 10s
- Native audio: yes
- Lipsync: no
- Character reference: no
- Open weights: yes
- Deprecated: no
- Best for: Alibaba prior-gen

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 56 |
| Motion realism | 60 |
| Character consistency | 57 |
| Temporal coherence | 55 |
| Lipsync accuracy | 54 |
| Native audio quality | 56 |
| Speed to first frame | 70 |
| Cost efficiency | 46 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.09 | 71s |
| fal.ai | $0.15 | 61s |
| RunningHub | $0.153 | 79s |
| Segmind | $0.188 | 68s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on Wan 2.6

## 30. LTX-2 Pro

- Page: https://vidbenchmark.com/model/ltx-2-pro
- Vendor: Lightricks
- Released: 2025-11-05
- Elo: 962 (-5 over 7 days)
- Price: $0.06/second
- Max resolution: 1080p
- Max duration: 6s
- Native audio: no
- Lipsync: no
- Character reference: no
- Open weights: yes
- Deprecated: no
- Best for: Open-weight legacy

### Scores (0-100)

| Axis | Score |
| --- | --- |
| Prompt adherence | 60 |
| Motion realism | 57 |
| Character consistency | 51 |
| Temporal coherence | 58 |
| Lipsync accuracy | 0 |
| Native audio quality | 0 |
| Speed to first frame | 71 |
| Cost efficiency | 57 |

### Providers

| Provider | Price/second | Latency |
| --- | --- | --- |
| Kie.ai | $0.036 | 64s |
| Novita | $0.053 | 74s |
| fal.ai | $0.06 | 54s |
| RunningHub | $0.061 | 71s |

### Sample prompts

- A street violinist plays in the rain while a crowd gathers, tested on LTX-2 Pro

## Head-to-head verdicts

### MiniMax H3 vs Seedance 2.0

- Page: https://vidbenchmark.com/vs/minimax-h3-vs-seedance-2
- Winner: MiniMax H3 (Elo 1268)
- Runner-up: Seedance 2.0 (Elo 1261)
- Cheaper: MiniMax H3 at $0.13/second, 14% below Seedance 2.0

MiniMax H3 wins the head-to-head on blind Elo (1268 vs 1261) and is the safer default for cinema quality + audio. MiniMax H3 is the cheaper option at $0.13 per second, which saves $1.26 per finished minute. Choose MiniMax H3 when render quality decides the job, and MiniMax H3 when volume and budget do. For dialogue scenes pick Seedance 2.0, which scores 93 on lipsync accuracy against 89.

### MiniMax H3 vs Gemini Omni Flash

- Page: https://vidbenchmark.com/vs/minimax-h3-vs-gemini-omni-flash
- Winner: MiniMax H3 (Elo 1268)
- Runner-up: Gemini Omni Flash (Elo 1254)
- Cheaper: Gemini Omni Flash at $0.1/second, 23% below MiniMax H3

MiniMax H3 wins the head-to-head on blind Elo (1268 vs 1254) and is the safer default for cinema quality + audio. Gemini Omni Flash is the cheaper option at $0.1 per second, which saves $1.80 per finished minute. Choose MiniMax H3 when render quality decides the job, and Gemini Omni Flash when volume and budget do. For dialogue scenes pick Gemini Omni Flash, which scores 93 on lipsync accuracy against 89.

### MiniMax H3 vs Grok Imagine 1.5

- Page: https://vidbenchmark.com/vs/minimax-h3-vs-grok-imagine-1-5
- Winner: MiniMax H3 (Elo 1268)
- Runner-up: Grok Imagine 1.5 (Elo 1241)
- Cheaper: MiniMax H3 at $0.13/second, 7% below Grok Imagine 1.5

MiniMax H3 wins the head-to-head on blind Elo (1268 vs 1241) and is the safer default for cinema quality + audio. MiniMax H3 is the cheaper option at $0.13 per second, which saves $0.60 per finished minute. Choose MiniMax H3 when render quality decides the job, and MiniMax H3 when volume and budget do. For dialogue scenes pick MiniMax H3, which scores 89 on lipsync accuracy against 84.

### MiniMax H3 vs HappyHorse 1.1

- Page: https://vidbenchmark.com/vs/minimax-h3-vs-happyhorse-1-1
- Winner: MiniMax H3 (Elo 1268)
- Runner-up: HappyHorse 1.1 (Elo 1226)
- Cheaper: MiniMax H3 at $0.13/second, 21% below HappyHorse 1.1

MiniMax H3 wins the head-to-head on blind Elo (1268 vs 1226) and is the safer default for cinema quality + audio. MiniMax H3 is the cheaper option at $0.13 per second, which saves $2.10 per finished minute. Choose MiniMax H3 when render quality decides the job, and MiniMax H3 when volume and budget do. For dialogue scenes pick MiniMax H3, which scores 89 on lipsync accuracy against 84.

### MiniMax H3 vs Wan 2.7

- Page: https://vidbenchmark.com/vs/minimax-h3-vs-wan-2-7
- Winner: MiniMax H3 (Elo 1268)
- Runner-up: Wan 2.7 (Elo 1215)
- Cheaper: MiniMax H3 at $0.13/second, 13% below Wan 2.7

MiniMax H3 wins the head-to-head on blind Elo (1268 vs 1215) and is the safer default for cinema quality + audio. MiniMax H3 is the cheaper option at $0.13 per second, which saves $1.20 per finished minute. Choose MiniMax H3 when render quality decides the job, and MiniMax H3 when volume and budget do. For dialogue scenes pick MiniMax H3, which scores 89 on lipsync accuracy against 86.

### MiniMax H3 vs SkyReels V4

- Page: https://vidbenchmark.com/vs/minimax-h3-vs-skyreels-v4
- Winner: MiniMax H3 (Elo 1268)
- Runner-up: SkyReels V4 (Elo 1206)
- Cheaper: MiniMax H3 at $0.13/second, 63% below SkyReels V4

MiniMax H3 wins the head-to-head on blind Elo (1268 vs 1206) and is the safer default for cinema quality + audio. MiniMax H3 is the cheaper option at $0.13 per second, which saves $13.20 per finished minute. Choose MiniMax H3 when render quality decides the job, and MiniMax H3 when volume and budget do. For dialogue scenes pick MiniMax H3, which scores 89 on lipsync accuracy against 84.

### MiniMax H3 vs Veo 3.1

- Page: https://vidbenchmark.com/vs/minimax-h3-vs-veo-3-1
- Winner: MiniMax H3 (Elo 1268)
- Runner-up: Veo 3.1 (Elo 1198)
- Cheaper: MiniMax H3 at $0.13/second, 68% below Veo 3.1

MiniMax H3 wins the head-to-head on blind Elo (1268 vs 1198) and is the safer default for cinema quality + audio. MiniMax H3 is the cheaper option at $0.13 per second, which saves $16.20 per finished minute. Choose MiniMax H3 when render quality decides the job, and MiniMax H3 when volume and budget do. For dialogue scenes pick MiniMax H3, which scores 89 on lipsync accuracy against 85.

### MiniMax H3 vs Grok Imagine

- Page: https://vidbenchmark.com/vs/minimax-h3-vs-grok-imagine
- Winner: MiniMax H3 (Elo 1268)
- Runner-up: Grok Imagine (Elo 1188)
- Cheaper: Grok Imagine at $0.07/second, 46% below MiniMax H3

MiniMax H3 wins the head-to-head on blind Elo (1268 vs 1188) and is the safer default for cinema quality + audio. Grok Imagine is the cheaper option at $0.07 per second, which saves $3.60 per finished minute. Choose MiniMax H3 when render quality decides the job, and Grok Imagine when volume and budget do. For dialogue scenes pick MiniMax H3, which scores 89 on lipsync accuracy against 83.

### MiniMax H3 vs Veo 3.1 Fast

- Page: https://vidbenchmark.com/vs/minimax-h3-vs-veo-3-1-fast
- Winner: MiniMax H3 (Elo 1268)
- Runner-up: Veo 3.1 Fast (Elo 1181)
- Cheaper: MiniMax H3 at $0.13/second, 13% below Veo 3.1 Fast

MiniMax H3 wins the head-to-head on blind Elo (1268 vs 1181) and is the safer default for cinema quality + audio. MiniMax H3 is the cheaper option at $0.13 per second, which saves $1.20 per finished minute. Choose MiniMax H3 when render quality decides the job, and MiniMax H3 when volume and budget do. For dialogue scenes pick MiniMax H3, which scores 89 on lipsync accuracy against 80.

### MiniMax H3 vs Kling 3.0 Pro 1080p

- Page: https://vidbenchmark.com/vs/minimax-h3-vs-kling-3-0-pro-1080p
- Winner: MiniMax H3 (Elo 1268)
- Runner-up: Kling 3.0 Pro 1080p (Elo 1173)
- Cheaper: MiniMax H3 at $0.13/second, 61% below Kling 3.0 Pro 1080p

MiniMax H3 wins the head-to-head on blind Elo (1268 vs 1173) and is the safer default for cinema quality + audio. MiniMax H3 is the cheaper option at $0.13 per second, which saves $12.36 per finished minute. Choose MiniMax H3 when render quality decides the job, and MiniMax H3 when volume and budget do. For dialogue scenes pick MiniMax H3, which scores 89 on lipsync accuracy against 78.

### MiniMax H3 vs PixVerse V6

- Page: https://vidbenchmark.com/vs/minimax-h3-vs-pixverse-v6
- Winner: MiniMax H3 (Elo 1268)
- Runner-up: PixVerse V6 (Elo 1165)
- Cheaper: PixVerse V6 at $0.115/second, 12% below MiniMax H3

MiniMax H3 wins the head-to-head on blind Elo (1268 vs 1165) and is the safer default for cinema quality + audio. PixVerse V6 is the cheaper option at $0.115 per second, which saves $0.90 per finished minute. Choose MiniMax H3 when render quality decides the job, and PixVerse V6 when volume and budget do. For dialogue scenes pick MiniMax H3, which scores 89 on lipsync accuracy against 76.

### Seedance 2.0 vs Gemini Omni Flash

- Page: https://vidbenchmark.com/vs/seedance-2-vs-gemini-omni-flash
- Winner: Seedance 2.0 (Elo 1261)
- Runner-up: Gemini Omni Flash (Elo 1254)
- Cheaper: Gemini Omni Flash at $0.1/second, 34% below Seedance 2.0

Seedance 2.0 wins the head-to-head on blind Elo (1261 vs 1254) and is the safer default for character-driven narrative. Gemini Omni Flash is the cheaper option at $0.1 per second, which saves $3.06 per finished minute. Choose Seedance 2.0 when render quality decides the job, and Gemini Omni Flash when volume and budget do. For dialogue scenes pick Seedance 2.0, which scores 93 on lipsync accuracy against 93.

### Seedance 2.0 vs Grok Imagine 1.5

- Page: https://vidbenchmark.com/vs/seedance-2-vs-grok-imagine-1-5
- Winner: Seedance 2.0 (Elo 1261)
- Runner-up: Grok Imagine 1.5 (Elo 1241)
- Cheaper: Grok Imagine 1.5 at $0.14/second, 7% below Seedance 2.0

Seedance 2.0 wins the head-to-head on blind Elo (1261 vs 1241) and is the safer default for character-driven narrative. Grok Imagine 1.5 is the cheaper option at $0.14 per second, which saves $0.66 per finished minute. Choose Seedance 2.0 when render quality decides the job, and Grok Imagine 1.5 when volume and budget do. For dialogue scenes pick Seedance 2.0, which scores 93 on lipsync accuracy against 84.

### Seedance 2.0 vs HappyHorse 1.1

- Page: https://vidbenchmark.com/vs/seedance-2-vs-happyhorse-1-1
- Winner: Seedance 2.0 (Elo 1261)
- Runner-up: HappyHorse 1.1 (Elo 1226)
- Cheaper: Seedance 2.0 at $0.151/second, 8% below HappyHorse 1.1

Seedance 2.0 wins the head-to-head on blind Elo (1261 vs 1226) and is the safer default for character-driven narrative. Seedance 2.0 is the cheaper option at $0.151 per second, which saves $0.84 per finished minute. Choose Seedance 2.0 when render quality decides the job, and Seedance 2.0 when volume and budget do. For dialogue scenes pick Seedance 2.0, which scores 93 on lipsync accuracy against 84.

### Seedance 2.0 vs Wan 2.7

- Page: https://vidbenchmark.com/vs/seedance-2-vs-wan-2-7
- Winner: Seedance 2.0 (Elo 1261)
- Runner-up: Wan 2.7 (Elo 1215)
- Cheaper: Wan 2.7 at $0.15/second, 1% below Seedance 2.0

Seedance 2.0 wins the head-to-head on blind Elo (1261 vs 1215) and is the safer default for character-driven narrative. Wan 2.7 is the cheaper option at $0.15 per second, which saves $0.06 per finished minute. Choose Seedance 2.0 when render quality decides the job, and Wan 2.7 when volume and budget do. For dialogue scenes pick Seedance 2.0, which scores 93 on lipsync accuracy against 86.

### Seedance 2.0 vs SkyReels V4

- Page: https://vidbenchmark.com/vs/seedance-2-vs-skyreels-v4
- Winner: Seedance 2.0 (Elo 1261)
- Runner-up: SkyReels V4 (Elo 1206)
- Cheaper: Seedance 2.0 at $0.151/second, 57% below SkyReels V4

Seedance 2.0 wins the head-to-head on blind Elo (1261 vs 1206) and is the safer default for character-driven narrative. Seedance 2.0 is the cheaper option at $0.151 per second, which saves $11.94 per finished minute. Choose Seedance 2.0 when render quality decides the job, and Seedance 2.0 when volume and budget do. For dialogue scenes pick Seedance 2.0, which scores 93 on lipsync accuracy against 84.

### Seedance 2.0 vs Veo 3.1

- Page: https://vidbenchmark.com/vs/seedance-2-vs-veo-3-1
- Winner: Seedance 2.0 (Elo 1261)
- Runner-up: Veo 3.1 (Elo 1198)
- Cheaper: Seedance 2.0 at $0.151/second, 62% below Veo 3.1

Seedance 2.0 wins the head-to-head on blind Elo (1261 vs 1198) and is the safer default for character-driven narrative. Seedance 2.0 is the cheaper option at $0.151 per second, which saves $14.94 per finished minute. Choose Seedance 2.0 when render quality decides the job, and Seedance 2.0 when volume and budget do. For dialogue scenes pick Seedance 2.0, which scores 93 on lipsync accuracy against 85.

### Seedance 2.0 vs Grok Imagine

- Page: https://vidbenchmark.com/vs/seedance-2-vs-grok-imagine
- Winner: Seedance 2.0 (Elo 1261)
- Runner-up: Grok Imagine (Elo 1188)
- Cheaper: Grok Imagine at $0.07/second, 54% below Seedance 2.0

Seedance 2.0 wins the head-to-head on blind Elo (1261 vs 1188) and is the safer default for character-driven narrative. Grok Imagine is the cheaper option at $0.07 per second, which saves $4.86 per finished minute. Choose Seedance 2.0 when render quality decides the job, and Grok Imagine when volume and budget do. For dialogue scenes pick Seedance 2.0, which scores 93 on lipsync accuracy against 83.

### Seedance 2.0 vs Veo 3.1 Fast

- Page: https://vidbenchmark.com/vs/seedance-2-vs-veo-3-1-fast
- Winner: Seedance 2.0 (Elo 1261)
- Runner-up: Veo 3.1 Fast (Elo 1181)
- Cheaper: Veo 3.1 Fast at $0.15/second, 1% below Seedance 2.0

Seedance 2.0 wins the head-to-head on blind Elo (1261 vs 1181) and is the safer default for character-driven narrative. Veo 3.1 Fast is the cheaper option at $0.15 per second, which saves $0.06 per finished minute. Choose Seedance 2.0 when render quality decides the job, and Veo 3.1 Fast when volume and budget do. For dialogue scenes pick Seedance 2.0, which scores 93 on lipsync accuracy against 80.

### Seedance 2.0 vs Kling 3.0 Pro 1080p

- Page: https://vidbenchmark.com/vs/seedance-2-vs-kling-3-0-pro-1080p
- Winner: Seedance 2.0 (Elo 1261)
- Runner-up: Kling 3.0 Pro 1080p (Elo 1173)
- Cheaper: Seedance 2.0 at $0.151/second, 55% below Kling 3.0 Pro 1080p

Seedance 2.0 wins the head-to-head on blind Elo (1261 vs 1173) and is the safer default for character-driven narrative. Seedance 2.0 is the cheaper option at $0.151 per second, which saves $11.10 per finished minute. Choose Seedance 2.0 when render quality decides the job, and Seedance 2.0 when volume and budget do. For dialogue scenes pick Seedance 2.0, which scores 93 on lipsync accuracy against 78.

### Seedance 2.0 vs PixVerse V6

- Page: https://vidbenchmark.com/vs/seedance-2-vs-pixverse-v6
- Winner: Seedance 2.0 (Elo 1261)
- Runner-up: PixVerse V6 (Elo 1165)
- Cheaper: PixVerse V6 at $0.115/second, 24% below Seedance 2.0

Seedance 2.0 wins the head-to-head on blind Elo (1261 vs 1165) and is the safer default for character-driven narrative. PixVerse V6 is the cheaper option at $0.115 per second, which saves $2.16 per finished minute. Choose Seedance 2.0 when render quality decides the job, and PixVerse V6 when volume and budget do. For dialogue scenes pick Seedance 2.0, which scores 93 on lipsync accuracy against 76.

### Gemini Omni Flash vs Grok Imagine 1.5

- Page: https://vidbenchmark.com/vs/gemini-omni-flash-vs-grok-imagine-1-5
- Winner: Gemini Omni Flash (Elo 1254)
- Runner-up: Grok Imagine 1.5 (Elo 1241)
- Cheaper: Gemini Omni Flash at $0.1/second, 29% below Grok Imagine 1.5

Gemini Omni Flash wins the head-to-head on blind Elo (1254 vs 1241) and is the safer default for fast prompt adherence. Gemini Omni Flash is the cheaper option at $0.1 per second, which saves $2.40 per finished minute. Choose Gemini Omni Flash when render quality decides the job, and Gemini Omni Flash when volume and budget do. For dialogue scenes pick Gemini Omni Flash, which scores 93 on lipsync accuracy against 84.

### Gemini Omni Flash vs HappyHorse 1.1

- Page: https://vidbenchmark.com/vs/gemini-omni-flash-vs-happyhorse-1-1
- Winner: Gemini Omni Flash (Elo 1254)
- Runner-up: HappyHorse 1.1 (Elo 1226)
- Cheaper: Gemini Omni Flash at $0.1/second, 39% below HappyHorse 1.1

Gemini Omni Flash wins the head-to-head on blind Elo (1254 vs 1226) and is the safer default for fast prompt adherence. Gemini Omni Flash is the cheaper option at $0.1 per second, which saves $3.90 per finished minute. Choose Gemini Omni Flash when render quality decides the job, and Gemini Omni Flash when volume and budget do. For dialogue scenes pick Gemini Omni Flash, which scores 93 on lipsync accuracy against 84.

### Gemini Omni Flash vs Wan 2.7

- Page: https://vidbenchmark.com/vs/gemini-omni-flash-vs-wan-2-7
- Winner: Gemini Omni Flash (Elo 1254)
- Runner-up: Wan 2.7 (Elo 1215)
- Cheaper: Gemini Omni Flash at $0.1/second, 33% below Wan 2.7

Gemini Omni Flash wins the head-to-head on blind Elo (1254 vs 1215) and is the safer default for fast prompt adherence. Gemini Omni Flash is the cheaper option at $0.1 per second, which saves $3.00 per finished minute. Choose Gemini Omni Flash when render quality decides the job, and Gemini Omni Flash when volume and budget do. For dialogue scenes pick Gemini Omni Flash, which scores 93 on lipsync accuracy against 86.

### Gemini Omni Flash vs SkyReels V4

- Page: https://vidbenchmark.com/vs/gemini-omni-flash-vs-skyreels-v4
- Winner: Gemini Omni Flash (Elo 1254)
- Runner-up: SkyReels V4 (Elo 1206)
- Cheaper: Gemini Omni Flash at $0.1/second, 71% below SkyReels V4

Gemini Omni Flash wins the head-to-head on blind Elo (1254 vs 1206) and is the safer default for fast prompt adherence. Gemini Omni Flash is the cheaper option at $0.1 per second, which saves $15.00 per finished minute. Choose Gemini Omni Flash when render quality decides the job, and Gemini Omni Flash when volume and budget do. For dialogue scenes pick Gemini Omni Flash, which scores 93 on lipsync accuracy against 84.

### Gemini Omni Flash vs Veo 3.1

- Page: https://vidbenchmark.com/vs/gemini-omni-flash-vs-veo-3-1
- Winner: Gemini Omni Flash (Elo 1254)
- Runner-up: Veo 3.1 (Elo 1198)
- Cheaper: Gemini Omni Flash at $0.1/second, 75% below Veo 3.1

Gemini Omni Flash wins the head-to-head on blind Elo (1254 vs 1198) and is the safer default for fast prompt adherence. Gemini Omni Flash is the cheaper option at $0.1 per second, which saves $18.00 per finished minute. Choose Gemini Omni Flash when render quality decides the job, and Gemini Omni Flash when volume and budget do. For dialogue scenes pick Gemini Omni Flash, which scores 93 on lipsync accuracy against 85.

### Gemini Omni Flash vs Grok Imagine

- Page: https://vidbenchmark.com/vs/gemini-omni-flash-vs-grok-imagine
- Winner: Gemini Omni Flash (Elo 1254)
- Runner-up: Grok Imagine (Elo 1188)
- Cheaper: Grok Imagine at $0.07/second, 30% below Gemini Omni Flash

Gemini Omni Flash wins the head-to-head on blind Elo (1254 vs 1188) and is the safer default for fast prompt adherence. Grok Imagine is the cheaper option at $0.07 per second, which saves $1.80 per finished minute. Choose Gemini Omni Flash when render quality decides the job, and Grok Imagine when volume and budget do. For dialogue scenes pick Gemini Omni Flash, which scores 93 on lipsync accuracy against 83.

### Gemini Omni Flash vs Veo 3.1 Fast

- Page: https://vidbenchmark.com/vs/gemini-omni-flash-vs-veo-3-1-fast
- Winner: Gemini Omni Flash (Elo 1254)
- Runner-up: Veo 3.1 Fast (Elo 1181)
- Cheaper: Gemini Omni Flash at $0.1/second, 33% below Veo 3.1 Fast

Gemini Omni Flash wins the head-to-head on blind Elo (1254 vs 1181) and is the safer default for fast prompt adherence. Gemini Omni Flash is the cheaper option at $0.1 per second, which saves $3.00 per finished minute. Choose Gemini Omni Flash when render quality decides the job, and Gemini Omni Flash when volume and budget do. For dialogue scenes pick Gemini Omni Flash, which scores 93 on lipsync accuracy against 80.

### Gemini Omni Flash vs Kling 3.0 Pro 1080p

- Page: https://vidbenchmark.com/vs/gemini-omni-flash-vs-kling-3-0-pro-1080p
- Winner: Gemini Omni Flash (Elo 1254)
- Runner-up: Kling 3.0 Pro 1080p (Elo 1173)
- Cheaper: Gemini Omni Flash at $0.1/second, 70% below Kling 3.0 Pro 1080p

Gemini Omni Flash wins the head-to-head on blind Elo (1254 vs 1173) and is the safer default for fast prompt adherence. Gemini Omni Flash is the cheaper option at $0.1 per second, which saves $14.16 per finished minute. Choose Gemini Omni Flash when render quality decides the job, and Gemini Omni Flash when volume and budget do. For dialogue scenes pick Gemini Omni Flash, which scores 93 on lipsync accuracy against 78.

### Gemini Omni Flash vs PixVerse V6

- Page: https://vidbenchmark.com/vs/gemini-omni-flash-vs-pixverse-v6
- Winner: Gemini Omni Flash (Elo 1254)
- Runner-up: PixVerse V6 (Elo 1165)
- Cheaper: Gemini Omni Flash at $0.1/second, 13% below PixVerse V6

Gemini Omni Flash wins the head-to-head on blind Elo (1254 vs 1165) and is the safer default for fast prompt adherence. Gemini Omni Flash is the cheaper option at $0.1 per second, which saves $0.90 per finished minute. Choose Gemini Omni Flash when render quality decides the job, and Gemini Omni Flash when volume and budget do. For dialogue scenes pick Gemini Omni Flash, which scores 93 on lipsync accuracy against 76.

### Grok Imagine 1.5 vs HappyHorse 1.1

- Page: https://vidbenchmark.com/vs/grok-imagine-1-5-vs-happyhorse-1-1
- Winner: Grok Imagine 1.5 (Elo 1241)
- Runner-up: HappyHorse 1.1 (Elo 1226)
- Cheaper: Grok Imagine 1.5 at $0.14/second, 15% below HappyHorse 1.1

Grok Imagine 1.5 wins the head-to-head on blind Elo (1241 vs 1226) and is the safer default for creative freedom. Grok Imagine 1.5 is the cheaper option at $0.14 per second, which saves $1.50 per finished minute. Choose Grok Imagine 1.5 when render quality decides the job, and Grok Imagine 1.5 when volume and budget do. For dialogue scenes pick Grok Imagine 1.5, which scores 84 on lipsync accuracy against 84.

### Grok Imagine 1.5 vs Wan 2.7

- Page: https://vidbenchmark.com/vs/grok-imagine-1-5-vs-wan-2-7
- Winner: Grok Imagine 1.5 (Elo 1241)
- Runner-up: Wan 2.7 (Elo 1215)
- Cheaper: Grok Imagine 1.5 at $0.14/second, 7% below Wan 2.7

Grok Imagine 1.5 wins the head-to-head on blind Elo (1241 vs 1215) and is the safer default for creative freedom. Grok Imagine 1.5 is the cheaper option at $0.14 per second, which saves $0.60 per finished minute. Choose Grok Imagine 1.5 when render quality decides the job, and Grok Imagine 1.5 when volume and budget do. For dialogue scenes pick Wan 2.7, which scores 86 on lipsync accuracy against 84.

### Grok Imagine 1.5 vs SkyReels V4

- Page: https://vidbenchmark.com/vs/grok-imagine-1-5-vs-skyreels-v4
- Winner: Grok Imagine 1.5 (Elo 1241)
- Runner-up: SkyReels V4 (Elo 1206)
- Cheaper: Grok Imagine 1.5 at $0.14/second, 60% below SkyReels V4

Grok Imagine 1.5 wins the head-to-head on blind Elo (1241 vs 1206) and is the safer default for creative freedom. Grok Imagine 1.5 is the cheaper option at $0.14 per second, which saves $12.60 per finished minute. Choose Grok Imagine 1.5 when render quality decides the job, and Grok Imagine 1.5 when volume and budget do. For dialogue scenes pick Grok Imagine 1.5, which scores 84 on lipsync accuracy against 84.

### Grok Imagine 1.5 vs Veo 3.1

- Page: https://vidbenchmark.com/vs/grok-imagine-1-5-vs-veo-3-1
- Winner: Grok Imagine 1.5 (Elo 1241)
- Runner-up: Veo 3.1 (Elo 1198)
- Cheaper: Grok Imagine 1.5 at $0.14/second, 65% below Veo 3.1

Grok Imagine 1.5 wins the head-to-head on blind Elo (1241 vs 1198) and is the safer default for creative freedom. Grok Imagine 1.5 is the cheaper option at $0.14 per second, which saves $15.60 per finished minute. Choose Grok Imagine 1.5 when render quality decides the job, and Grok Imagine 1.5 when volume and budget do. For dialogue scenes pick Veo 3.1, which scores 85 on lipsync accuracy against 84.

### Grok Imagine 1.5 vs Grok Imagine

- Page: https://vidbenchmark.com/vs/grok-imagine-1-5-vs-grok-imagine
- Winner: Grok Imagine 1.5 (Elo 1241)
- Runner-up: Grok Imagine (Elo 1188)
- Cheaper: Grok Imagine at $0.07/second, 50% below Grok Imagine 1.5

Grok Imagine 1.5 wins the head-to-head on blind Elo (1241 vs 1188) and is the safer default for creative freedom. Grok Imagine is the cheaper option at $0.07 per second, which saves $4.20 per finished minute. Choose Grok Imagine 1.5 when render quality decides the job, and Grok Imagine when volume and budget do. For dialogue scenes pick Grok Imagine 1.5, which scores 84 on lipsync accuracy against 83.

### Grok Imagine 1.5 vs Veo 3.1 Fast

- Page: https://vidbenchmark.com/vs/grok-imagine-1-5-vs-veo-3-1-fast
- Winner: Grok Imagine 1.5 (Elo 1241)
- Runner-up: Veo 3.1 Fast (Elo 1181)
- Cheaper: Grok Imagine 1.5 at $0.14/second, 7% below Veo 3.1 Fast

Grok Imagine 1.5 wins the head-to-head on blind Elo (1241 vs 1181) and is the safer default for creative freedom. Grok Imagine 1.5 is the cheaper option at $0.14 per second, which saves $0.60 per finished minute. Choose Grok Imagine 1.5 when render quality decides the job, and Grok Imagine 1.5 when volume and budget do. For dialogue scenes pick Grok Imagine 1.5, which scores 84 on lipsync accuracy against 80.

### Grok Imagine 1.5 vs Kling 3.0 Pro 1080p

- Page: https://vidbenchmark.com/vs/grok-imagine-1-5-vs-kling-3-0-pro-1080p
- Winner: Grok Imagine 1.5 (Elo 1241)
- Runner-up: Kling 3.0 Pro 1080p (Elo 1173)
- Cheaper: Grok Imagine 1.5 at $0.14/second, 58% below Kling 3.0 Pro 1080p

Grok Imagine 1.5 wins the head-to-head on blind Elo (1241 vs 1173) and is the safer default for creative freedom. Grok Imagine 1.5 is the cheaper option at $0.14 per second, which saves $11.76 per finished minute. Choose Grok Imagine 1.5 when render quality decides the job, and Grok Imagine 1.5 when volume and budget do. For dialogue scenes pick Grok Imagine 1.5, which scores 84 on lipsync accuracy against 78.

### Grok Imagine 1.5 vs PixVerse V6

- Page: https://vidbenchmark.com/vs/grok-imagine-1-5-vs-pixverse-v6
- Winner: Grok Imagine 1.5 (Elo 1241)
- Runner-up: PixVerse V6 (Elo 1165)
- Cheaper: PixVerse V6 at $0.115/second, 18% below Grok Imagine 1.5

Grok Imagine 1.5 wins the head-to-head on blind Elo (1241 vs 1165) and is the safer default for creative freedom. PixVerse V6 is the cheaper option at $0.115 per second, which saves $1.50 per finished minute. Choose Grok Imagine 1.5 when render quality decides the job, and PixVerse V6 when volume and budget do. For dialogue scenes pick Grok Imagine 1.5, which scores 84 on lipsync accuracy against 76.

### HappyHorse 1.1 vs Wan 2.7

- Page: https://vidbenchmark.com/vs/happyhorse-1-1-vs-wan-2-7
- Winner: HappyHorse 1.1 (Elo 1226)
- Runner-up: Wan 2.7 (Elo 1215)
- Cheaper: Wan 2.7 at $0.15/second, 9% below HappyHorse 1.1

HappyHorse 1.1 wins the head-to-head on blind Elo (1226 vs 1215) and is the safer default for natural movement. Wan 2.7 is the cheaper option at $0.15 per second, which saves $0.90 per finished minute. Choose HappyHorse 1.1 when render quality decides the job, and Wan 2.7 when volume and budget do. For dialogue scenes pick Wan 2.7, which scores 86 on lipsync accuracy against 84.

### HappyHorse 1.1 vs SkyReels V4

- Page: https://vidbenchmark.com/vs/happyhorse-1-1-vs-skyreels-v4
- Winner: HappyHorse 1.1 (Elo 1226)
- Runner-up: SkyReels V4 (Elo 1206)
- Cheaper: HappyHorse 1.1 at $0.165/second, 53% below SkyReels V4

HappyHorse 1.1 wins the head-to-head on blind Elo (1226 vs 1206) and is the safer default for natural movement. HappyHorse 1.1 is the cheaper option at $0.165 per second, which saves $11.10 per finished minute. Choose HappyHorse 1.1 when render quality decides the job, and HappyHorse 1.1 when volume and budget do. For dialogue scenes pick HappyHorse 1.1, which scores 84 on lipsync accuracy against 84.

### HappyHorse 1.1 vs Veo 3.1

- Page: https://vidbenchmark.com/vs/happyhorse-1-1-vs-veo-3-1
- Winner: HappyHorse 1.1 (Elo 1226)
- Runner-up: Veo 3.1 (Elo 1198)
- Cheaper: HappyHorse 1.1 at $0.165/second, 59% below Veo 3.1

HappyHorse 1.1 wins the head-to-head on blind Elo (1226 vs 1198) and is the safer default for natural movement. HappyHorse 1.1 is the cheaper option at $0.165 per second, which saves $14.10 per finished minute. Choose HappyHorse 1.1 when render quality decides the job, and HappyHorse 1.1 when volume and budget do. For dialogue scenes pick Veo 3.1, which scores 85 on lipsync accuracy against 84.

### HappyHorse 1.1 vs Grok Imagine

- Page: https://vidbenchmark.com/vs/happyhorse-1-1-vs-grok-imagine
- Winner: HappyHorse 1.1 (Elo 1226)
- Runner-up: Grok Imagine (Elo 1188)
- Cheaper: Grok Imagine at $0.07/second, 58% below HappyHorse 1.1

HappyHorse 1.1 wins the head-to-head on blind Elo (1226 vs 1188) and is the safer default for natural movement. Grok Imagine is the cheaper option at $0.07 per second, which saves $5.70 per finished minute. Choose HappyHorse 1.1 when render quality decides the job, and Grok Imagine when volume and budget do. For dialogue scenes pick HappyHorse 1.1, which scores 84 on lipsync accuracy against 83.

### HappyHorse 1.1 vs Veo 3.1 Fast

- Page: https://vidbenchmark.com/vs/happyhorse-1-1-vs-veo-3-1-fast
- Winner: HappyHorse 1.1 (Elo 1226)
- Runner-up: Veo 3.1 Fast (Elo 1181)
- Cheaper: Veo 3.1 Fast at $0.15/second, 9% below HappyHorse 1.1

HappyHorse 1.1 wins the head-to-head on blind Elo (1226 vs 1181) and is the safer default for natural movement. Veo 3.1 Fast is the cheaper option at $0.15 per second, which saves $0.90 per finished minute. Choose HappyHorse 1.1 when render quality decides the job, and Veo 3.1 Fast when volume and budget do. For dialogue scenes pick HappyHorse 1.1, which scores 84 on lipsync accuracy against 80.

### HappyHorse 1.1 vs Kling 3.0 Pro 1080p

- Page: https://vidbenchmark.com/vs/happyhorse-1-1-vs-kling-3-0-pro-1080p
- Winner: HappyHorse 1.1 (Elo 1226)
- Runner-up: Kling 3.0 Pro 1080p (Elo 1173)
- Cheaper: HappyHorse 1.1 at $0.165/second, 51% below Kling 3.0 Pro 1080p

HappyHorse 1.1 wins the head-to-head on blind Elo (1226 vs 1173) and is the safer default for natural movement. HappyHorse 1.1 is the cheaper option at $0.165 per second, which saves $10.26 per finished minute. Choose HappyHorse 1.1 when render quality decides the job, and HappyHorse 1.1 when volume and budget do. For dialogue scenes pick HappyHorse 1.1, which scores 84 on lipsync accuracy against 78.

### HappyHorse 1.1 vs PixVerse V6

- Page: https://vidbenchmark.com/vs/happyhorse-1-1-vs-pixverse-v6
- Winner: HappyHorse 1.1 (Elo 1226)
- Runner-up: PixVerse V6 (Elo 1165)
- Cheaper: PixVerse V6 at $0.115/second, 30% below HappyHorse 1.1

HappyHorse 1.1 wins the head-to-head on blind Elo (1226 vs 1165) and is the safer default for natural movement. PixVerse V6 is the cheaper option at $0.115 per second, which saves $3.00 per finished minute. Choose HappyHorse 1.1 when render quality decides the job, and PixVerse V6 when volume and budget do. For dialogue scenes pick HappyHorse 1.1, which scores 84 on lipsync accuracy against 76.

### Wan 2.7 vs SkyReels V4

- Page: https://vidbenchmark.com/vs/wan-2-7-vs-skyreels-v4
- Winner: Wan 2.7 (Elo 1215)
- Runner-up: SkyReels V4 (Elo 1206)
- Cheaper: Wan 2.7 at $0.15/second, 57% below SkyReels V4

Wan 2.7 wins the head-to-head on blind Elo (1215 vs 1206) and is the safer default for best open-weight w/ audio. Wan 2.7 is the cheaper option at $0.15 per second, which saves $12.00 per finished minute. Choose Wan 2.7 when render quality decides the job, and Wan 2.7 when volume and budget do. For dialogue scenes pick Wan 2.7, which scores 86 on lipsync accuracy against 84.

### Wan 2.7 vs Veo 3.1

- Page: https://vidbenchmark.com/vs/wan-2-7-vs-veo-3-1
- Winner: Wan 2.7 (Elo 1215)
- Runner-up: Veo 3.1 (Elo 1198)
- Cheaper: Wan 2.7 at $0.15/second, 63% below Veo 3.1

Wan 2.7 wins the head-to-head on blind Elo (1215 vs 1198) and is the safer default for best open-weight w/ audio. Wan 2.7 is the cheaper option at $0.15 per second, which saves $15.00 per finished minute. Choose Wan 2.7 when render quality decides the job, and Wan 2.7 when volume and budget do. For dialogue scenes pick Wan 2.7, which scores 86 on lipsync accuracy against 85.

### Wan 2.7 vs Grok Imagine

- Page: https://vidbenchmark.com/vs/wan-2-7-vs-grok-imagine
- Winner: Wan 2.7 (Elo 1215)
- Runner-up: Grok Imagine (Elo 1188)
- Cheaper: Grok Imagine at $0.07/second, 53% below Wan 2.7

Wan 2.7 wins the head-to-head on blind Elo (1215 vs 1188) and is the safer default for best open-weight w/ audio. Grok Imagine is the cheaper option at $0.07 per second, which saves $4.80 per finished minute. Choose Wan 2.7 when render quality decides the job, and Grok Imagine when volume and budget do. For dialogue scenes pick Wan 2.7, which scores 86 on lipsync accuracy against 83.

### Wan 2.7 vs Veo 3.1 Fast

- Page: https://vidbenchmark.com/vs/wan-2-7-vs-veo-3-1-fast
- Winner: Wan 2.7 (Elo 1215)
- Runner-up: Veo 3.1 Fast (Elo 1181)
- Cheaper: Wan 2.7 at $0.15/second, 0% below Veo 3.1 Fast

Wan 2.7 wins the head-to-head on blind Elo (1215 vs 1181) and is the safer default for best open-weight w/ audio. Wan 2.7 is the cheaper option at $0.15 per second, which saves $0.00 per finished minute. Choose Wan 2.7 when render quality decides the job, and Wan 2.7 when volume and budget do. For dialogue scenes pick Wan 2.7, which scores 86 on lipsync accuracy against 80.

### Wan 2.7 vs Kling 3.0 Pro 1080p

- Page: https://vidbenchmark.com/vs/wan-2-7-vs-kling-3-0-pro-1080p
- Winner: Wan 2.7 (Elo 1215)
- Runner-up: Kling 3.0 Pro 1080p (Elo 1173)
- Cheaper: Wan 2.7 at $0.15/second, 55% below Kling 3.0 Pro 1080p

Wan 2.7 wins the head-to-head on blind Elo (1215 vs 1173) and is the safer default for best open-weight w/ audio. Wan 2.7 is the cheaper option at $0.15 per second, which saves $11.16 per finished minute. Choose Wan 2.7 when render quality decides the job, and Wan 2.7 when volume and budget do. For dialogue scenes pick Wan 2.7, which scores 86 on lipsync accuracy against 78.

### Wan 2.7 vs PixVerse V6

- Page: https://vidbenchmark.com/vs/wan-2-7-vs-pixverse-v6
- Winner: Wan 2.7 (Elo 1215)
- Runner-up: PixVerse V6 (Elo 1165)
- Cheaper: PixVerse V6 at $0.115/second, 23% below Wan 2.7

Wan 2.7 wins the head-to-head on blind Elo (1215 vs 1165) and is the safer default for best open-weight w/ audio. PixVerse V6 is the cheaper option at $0.115 per second, which saves $2.10 per finished minute. Choose Wan 2.7 when render quality decides the job, and PixVerse V6 when volume and budget do. For dialogue scenes pick Wan 2.7, which scores 86 on lipsync accuracy against 76.

### SkyReels V4 vs Veo 3.1

- Page: https://vidbenchmark.com/vs/skyreels-v4-vs-veo-3-1
- Winner: SkyReels V4 (Elo 1206)
- Runner-up: Veo 3.1 (Elo 1198)
- Cheaper: SkyReels V4 at $0.35/second, 13% below Veo 3.1

SkyReels V4 wins the head-to-head on blind Elo (1206 vs 1198) and is the safer default for long-form storytelling. SkyReels V4 is the cheaper option at $0.35 per second, which saves $3.00 per finished minute. Choose SkyReels V4 when render quality decides the job, and SkyReels V4 when volume and budget do. For dialogue scenes pick Veo 3.1, which scores 85 on lipsync accuracy against 84.

### SkyReels V4 vs Grok Imagine

- Page: https://vidbenchmark.com/vs/skyreels-v4-vs-grok-imagine
- Winner: SkyReels V4 (Elo 1206)
- Runner-up: Grok Imagine (Elo 1188)
- Cheaper: Grok Imagine at $0.07/second, 80% below SkyReels V4

SkyReels V4 wins the head-to-head on blind Elo (1206 vs 1188) and is the safer default for long-form storytelling. Grok Imagine is the cheaper option at $0.07 per second, which saves $16.80 per finished minute. Choose SkyReels V4 when render quality decides the job, and Grok Imagine when volume and budget do. For dialogue scenes pick SkyReels V4, which scores 84 on lipsync accuracy against 83.

### SkyReels V4 vs Veo 3.1 Fast

- Page: https://vidbenchmark.com/vs/skyreels-v4-vs-veo-3-1-fast
- Winner: SkyReels V4 (Elo 1206)
- Runner-up: Veo 3.1 Fast (Elo 1181)
- Cheaper: Veo 3.1 Fast at $0.15/second, 57% below SkyReels V4

SkyReels V4 wins the head-to-head on blind Elo (1206 vs 1181) and is the safer default for long-form storytelling. Veo 3.1 Fast is the cheaper option at $0.15 per second, which saves $12.00 per finished minute. Choose SkyReels V4 when render quality decides the job, and Veo 3.1 Fast when volume and budget do. For dialogue scenes pick SkyReels V4, which scores 84 on lipsync accuracy against 80.

### SkyReels V4 vs Kling 3.0 Pro 1080p

- Page: https://vidbenchmark.com/vs/skyreels-v4-vs-kling-3-0-pro-1080p
- Winner: SkyReels V4 (Elo 1206)
- Runner-up: Kling 3.0 Pro 1080p (Elo 1173)
- Cheaper: Kling 3.0 Pro 1080p at $0.336/second, 4% below SkyReels V4

SkyReels V4 wins the head-to-head on blind Elo (1206 vs 1173) and is the safer default for long-form storytelling. Kling 3.0 Pro 1080p is the cheaper option at $0.336 per second, which saves $0.84 per finished minute. Choose SkyReels V4 when render quality decides the job, and Kling 3.0 Pro 1080p when volume and budget do. For dialogue scenes pick SkyReels V4, which scores 84 on lipsync accuracy against 78.

### SkyReels V4 vs PixVerse V6

- Page: https://vidbenchmark.com/vs/skyreels-v4-vs-pixverse-v6
- Winner: SkyReels V4 (Elo 1206)
- Runner-up: PixVerse V6 (Elo 1165)
- Cheaper: PixVerse V6 at $0.115/second, 67% below SkyReels V4

SkyReels V4 wins the head-to-head on blind Elo (1206 vs 1165) and is the safer default for long-form storytelling. PixVerse V6 is the cheaper option at $0.115 per second, which saves $14.10 per finished minute. Choose SkyReels V4 when render quality decides the job, and PixVerse V6 when volume and budget do. For dialogue scenes pick SkyReels V4, which scores 84 on lipsync accuracy against 76.

### Veo 3.1 vs Grok Imagine

- Page: https://vidbenchmark.com/vs/veo-3-1-vs-grok-imagine
- Winner: Veo 3.1 (Elo 1198)
- Runner-up: Grok Imagine (Elo 1188)
- Cheaper: Grok Imagine at $0.07/second, 83% below Veo 3.1

Veo 3.1 wins the head-to-head on blind Elo (1198 vs 1188) and is the safer default for prompt adherence king. Grok Imagine is the cheaper option at $0.07 per second, which saves $19.80 per finished minute. Choose Veo 3.1 when render quality decides the job, and Grok Imagine when volume and budget do. For dialogue scenes pick Veo 3.1, which scores 85 on lipsync accuracy against 83.

### Veo 3.1 vs Veo 3.1 Fast

- Page: https://vidbenchmark.com/vs/veo-3-1-vs-veo-3-1-fast
- Winner: Veo 3.1 (Elo 1198)
- Runner-up: Veo 3.1 Fast (Elo 1181)
- Cheaper: Veo 3.1 Fast at $0.15/second, 63% below Veo 3.1

Veo 3.1 wins the head-to-head on blind Elo (1198 vs 1181) and is the safer default for prompt adherence king. Veo 3.1 Fast is the cheaper option at $0.15 per second, which saves $15.00 per finished minute. Choose Veo 3.1 when render quality decides the job, and Veo 3.1 Fast when volume and budget do. For dialogue scenes pick Veo 3.1, which scores 85 on lipsync accuracy against 80.

### Veo 3.1 vs Kling 3.0 Pro 1080p

- Page: https://vidbenchmark.com/vs/veo-3-1-vs-kling-3-0-pro-1080p
- Winner: Veo 3.1 (Elo 1198)
- Runner-up: Kling 3.0 Pro 1080p (Elo 1173)
- Cheaper: Kling 3.0 Pro 1080p at $0.336/second, 16% below Veo 3.1

Veo 3.1 wins the head-to-head on blind Elo (1198 vs 1173) and is the safer default for prompt adherence king. Kling 3.0 Pro 1080p is the cheaper option at $0.336 per second, which saves $3.84 per finished minute. Choose Veo 3.1 when render quality decides the job, and Kling 3.0 Pro 1080p when volume and budget do. For dialogue scenes pick Veo 3.1, which scores 85 on lipsync accuracy against 78.

### Veo 3.1 vs PixVerse V6

- Page: https://vidbenchmark.com/vs/veo-3-1-vs-pixverse-v6
- Winner: Veo 3.1 (Elo 1198)
- Runner-up: PixVerse V6 (Elo 1165)
- Cheaper: PixVerse V6 at $0.115/second, 71% below Veo 3.1

Veo 3.1 wins the head-to-head on blind Elo (1198 vs 1165) and is the safer default for prompt adherence king. PixVerse V6 is the cheaper option at $0.115 per second, which saves $17.10 per finished minute. Choose Veo 3.1 when render quality decides the job, and PixVerse V6 when volume and budget do. For dialogue scenes pick Veo 3.1, which scores 85 on lipsync accuracy against 76.

### Grok Imagine vs Veo 3.1 Fast

- Page: https://vidbenchmark.com/vs/grok-imagine-vs-veo-3-1-fast
- Winner: Grok Imagine (Elo 1188)
- Runner-up: Veo 3.1 Fast (Elo 1181)
- Cheaper: Grok Imagine at $0.07/second, 53% below Veo 3.1 Fast

Grok Imagine wins the head-to-head on blind Elo (1188 vs 1181) and is the safer default for grok v1. Grok Imagine is the cheaper option at $0.07 per second, which saves $4.80 per finished minute. Choose Grok Imagine when render quality decides the job, and Grok Imagine when volume and budget do. For dialogue scenes pick Grok Imagine, which scores 83 on lipsync accuracy against 80.

### Grok Imagine vs Kling 3.0 Pro 1080p

- Page: https://vidbenchmark.com/vs/grok-imagine-vs-kling-3-0-pro-1080p
- Winner: Grok Imagine (Elo 1188)
- Runner-up: Kling 3.0 Pro 1080p (Elo 1173)
- Cheaper: Grok Imagine at $0.07/second, 79% below Kling 3.0 Pro 1080p

Grok Imagine wins the head-to-head on blind Elo (1188 vs 1173) and is the safer default for grok v1. Grok Imagine is the cheaper option at $0.07 per second, which saves $15.96 per finished minute. Choose Grok Imagine when render quality decides the job, and Grok Imagine when volume and budget do. For dialogue scenes pick Grok Imagine, which scores 83 on lipsync accuracy against 78.

### Grok Imagine vs PixVerse V6

- Page: https://vidbenchmark.com/vs/grok-imagine-vs-pixverse-v6
- Winner: Grok Imagine (Elo 1188)
- Runner-up: PixVerse V6 (Elo 1165)
- Cheaper: Grok Imagine at $0.07/second, 39% below PixVerse V6

Grok Imagine wins the head-to-head on blind Elo (1188 vs 1165) and is the safer default for grok v1. Grok Imagine is the cheaper option at $0.07 per second, which saves $2.70 per finished minute. Choose Grok Imagine when render quality decides the job, and Grok Imagine when volume and budget do. For dialogue scenes pick Grok Imagine, which scores 83 on lipsync accuracy against 76.

### Veo 3.1 Fast vs Kling 3.0 Pro 1080p

- Page: https://vidbenchmark.com/vs/veo-3-1-fast-vs-kling-3-0-pro-1080p
- Winner: Veo 3.1 Fast (Elo 1181)
- Runner-up: Kling 3.0 Pro 1080p (Elo 1173)
- Cheaper: Veo 3.1 Fast at $0.15/second, 55% below Kling 3.0 Pro 1080p

Veo 3.1 Fast wins the head-to-head on blind Elo (1181 vs 1173) and is the safer default for cheap google option. Veo 3.1 Fast is the cheaper option at $0.15 per second, which saves $11.16 per finished minute. Choose Veo 3.1 Fast when render quality decides the job, and Veo 3.1 Fast when volume and budget do. For dialogue scenes pick Veo 3.1 Fast, which scores 80 on lipsync accuracy against 78.

### Veo 3.1 Fast vs PixVerse V6

- Page: https://vidbenchmark.com/vs/veo-3-1-fast-vs-pixverse-v6
- Winner: Veo 3.1 Fast (Elo 1181)
- Runner-up: PixVerse V6 (Elo 1165)
- Cheaper: PixVerse V6 at $0.115/second, 23% below Veo 3.1 Fast

Veo 3.1 Fast wins the head-to-head on blind Elo (1181 vs 1165) and is the safer default for cheap google option. PixVerse V6 is the cheaper option at $0.115 per second, which saves $2.10 per finished minute. Choose Veo 3.1 Fast when render quality decides the job, and PixVerse V6 when volume and budget do. For dialogue scenes pick Veo 3.1 Fast, which scores 80 on lipsync accuracy against 76.

### Kling 3.0 Pro 1080p vs PixVerse V6

- Page: https://vidbenchmark.com/vs/kling-3-0-pro-1080p-vs-pixverse-v6
- Winner: Kling 3.0 Pro 1080p (Elo 1173)
- Runner-up: PixVerse V6 (Elo 1165)
- Cheaper: PixVerse V6 at $0.115/second, 66% below Kling 3.0 Pro 1080p

Kling 3.0 Pro 1080p wins the head-to-head on blind Elo (1173 vs 1165) and is the safer default for professional kling tier. PixVerse V6 is the cheaper option at $0.115 per second, which saves $13.26 per finished minute. Choose Kling 3.0 Pro 1080p when render quality decides the job, and PixVerse V6 when volume and budget do. For dialogue scenes pick Kling 3.0 Pro 1080p, which scores 78 on lipsync accuracy against 76.

### PixVerse V6 vs Kling 3.0 Standard 720p

- Page: https://vidbenchmark.com/vs/pixverse-v6-vs-kling-3-0-standard-720p
- Winner: PixVerse V6 (Elo 1165)
- Runner-up: Kling 3.0 Standard 720p (Elo 1156)
- Cheaper: PixVerse V6 at $0.115/second, 56% below Kling 3.0 Standard 720p

PixVerse V6 wins the head-to-head on blind Elo (1165 vs 1156) and is the safer default for anime + stylized. PixVerse V6 is the cheaper option at $0.115 per second, which saves $8.70 per finished minute. Choose PixVerse V6 when render quality decides the job, and PixVerse V6 when volume and budget do. For dialogue scenes pick PixVerse V6, which scores 76 on lipsync accuracy against 72.

### Kling 3.0 Standard 720p vs Kling 3 Omni 1080p

- Page: https://vidbenchmark.com/vs/kling-3-0-standard-720p-vs-kling-3-omni-1080p
- Winner: Kling 3.0 Standard 720p (Elo 1156)
- Runner-up: Kling 3 Omni 1080p (Elo 1148)
- Cheaper: Kling 3.0 Standard 720p at $0.26/second, 7% below Kling 3 Omni 1080p

Kling 3.0 Standard 720p wins the head-to-head on blind Elo (1156 vs 1148) and is the safer default for kling economy. Kling 3.0 Standard 720p is the cheaper option at $0.26 per second, which saves $1.20 per finished minute. Choose Kling 3.0 Standard 720p when render quality decides the job, and Kling 3.0 Standard 720p when volume and budget do. For dialogue scenes pick Kling 3.0 Standard 720p, which scores 72 on lipsync accuracy against 71.

### Kling 3 Omni 1080p vs Vidu Q3 Pro

- Page: https://vidbenchmark.com/vs/kling-3-omni-1080p-vs-vidu-q3-pro
- Winner: Kling 3 Omni 1080p (Elo 1148)
- Runner-up: Vidu Q3 Pro (Elo 1141)
- Cheaper: Vidu Q3 Pro at $0.16/second, 43% below Kling 3 Omni 1080p

Kling 3 Omni 1080p wins the head-to-head on blind Elo (1148 vs 1141) and is the safer default for voice-driven multi-scene. Vidu Q3 Pro is the cheaper option at $0.16 per second, which saves $7.20 per finished minute. Choose Kling 3 Omni 1080p when render quality decides the job, and Vidu Q3 Pro when volume and budget do. For dialogue scenes pick Vidu Q3 Pro, which scores 72 on lipsync accuracy against 71.

### Vidu Q3 Pro vs Kling 3 Omni 720p Std

- Page: https://vidbenchmark.com/vs/vidu-q3-pro-vs-kling-3-omni-720p-std
- Winner: Vidu Q3 Pro (Elo 1141)
- Runner-up: Kling 3 Omni 720p Std (Elo 1130)
- Cheaper: Vidu Q3 Pro at $0.16/second, 29% below Kling 3 Omni 720p Std

Vidu Q3 Pro wins the head-to-head on blind Elo (1141 vs 1130) and is the safer default for 16s length record. Vidu Q3 Pro is the cheaper option at $0.16 per second, which saves $3.84 per finished minute. Choose Vidu Q3 Pro when render quality decides the job, and Vidu Q3 Pro when volume and budget do. For dialogue scenes pick Vidu Q3 Pro, which scores 72 on lipsync accuracy against 70.

### Kling 3 Omni 720p Std vs Runway Gen-4 Turbo

- Page: https://vidbenchmark.com/vs/kling-3-omni-720p-std-vs-runway-gen-4
- Winner: Kling 3 Omni 720p Std (Elo 1130)
- Runner-up: Runway Gen-4 Turbo (Elo 1122)
- Cheaper: Runway Gen-4 Turbo at $0.18/second, 20% below Kling 3 Omni 720p Std

Kling 3 Omni 720p Std wins the head-to-head on blind Elo (1130 vs 1122) and is the safer default for kling voice-driven cheap. Runway Gen-4 Turbo is the cheaper option at $0.18 per second, which saves $2.64 per finished minute. Choose Kling 3 Omni 720p Std when render quality decides the job, and Runway Gen-4 Turbo when volume and budget do. For dialogue scenes pick Kling 3 Omni 720p Std, which scores 70 on lipsync accuracy against 0.

### Runway Gen-4 Turbo vs Sora 2 Pro

- Page: https://vidbenchmark.com/vs/runway-gen-4-vs-sora-2-pro
- Winner: Runway Gen-4 Turbo (Elo 1122)
- Runner-up: Sora 2 Pro (Elo 1113)
- Cheaper: Runway Gen-4 Turbo at $0.18/second, 64% below Sora 2 Pro

Runway Gen-4 Turbo wins the head-to-head on blind Elo (1122 vs 1113) and is the safer default for character consistency + editing. Runway Gen-4 Turbo is the cheaper option at $0.18 per second, which saves $19.20 per finished minute. Choose Runway Gen-4 Turbo when render quality decides the job, and Runway Gen-4 Turbo when volume and budget do. For dialogue scenes pick Sora 2 Pro, which scores 73 on lipsync accuracy against 0.

### Sora 2 Pro vs Sora 2

- Page: https://vidbenchmark.com/vs/sora-2-pro-vs-sora-2
- Winner: Sora 2 Pro (Elo 1113)
- Runner-up: Sora 2 (Elo 1098)
- Cheaper: Sora 2 at $0.1/second, 80% below Sora 2 Pro

Sora 2 Pro wins the head-to-head on blind Elo (1113 vs 1098) and is the safer default for deprecated sep 2026. Sora 2 is the cheaper option at $0.1 per second, which saves $24.00 per finished minute. Choose Sora 2 Pro when render quality decides the job, and Sora 2 when volume and budget do. For dialogue scenes pick Sora 2 Pro, which scores 73 on lipsync accuracy against 64.

### Sora 2 vs Kling 2.6 Pro

- Page: https://vidbenchmark.com/vs/sora-2-vs-kling-2-6-pro
- Winner: Sora 2 (Elo 1098)
- Runner-up: Kling 2.6 Pro (Elo 1090)
- Cheaper: Sora 2 at $0.1/second, 29% below Kling 2.6 Pro

Sora 2 wins the head-to-head on blind Elo (1098 vs 1090) and is the safer default for deprecated sep 2026. Sora 2 is the cheaper option at $0.1 per second, which saves $2.40 per finished minute. Choose Sora 2 when render quality decides the job, and Sora 2 when volume and budget do. For dialogue scenes pick Kling 2.6 Pro, which scores 69 on lipsync accuracy against 64.

### Kling 2.6 Pro vs Seedance 1.5 Pro

- Page: https://vidbenchmark.com/vs/kling-2-6-pro-vs-seedance-1-5-pro
- Winner: Kling 2.6 Pro (Elo 1090)
- Runner-up: Seedance 1.5 Pro (Elo 1081)
- Cheaper: Kling 2.6 Pro at $0.14/second, 29% below Seedance 1.5 Pro

Kling 2.6 Pro wins the head-to-head on blind Elo (1090 vs 1081) and is the safer default for prior-gen kling. Kling 2.6 Pro is the cheaper option at $0.14 per second, which saves $3.48 per finished minute. Choose Kling 2.6 Pro when render quality decides the job, and Kling 2.6 Pro when volume and budget do. For dialogue scenes pick Kling 2.6 Pro, which scores 69 on lipsync accuracy against 0.

### Seedance 1.5 Pro vs Cosmos3-Super-Image2Video

- Page: https://vidbenchmark.com/vs/seedance-1-5-pro-vs-cosmos3
- Winner: Seedance 1.5 Pro (Elo 1081)
- Runner-up: Cosmos3-Super-Image2Video (Elo 1070)
- Cheaper: Cosmos3-Super-Image2Video at $0.05/second, 75% below Seedance 1.5 Pro

Seedance 1.5 Pro wins the head-to-head on blind Elo (1081 vs 1070) and is the safer default for prior-gen seedance. Cosmos3-Super-Image2Video is the cheaper option at $0.05 per second, which saves $8.88 per finished minute. Choose Seedance 1.5 Pro when render quality decides the job, and Cosmos3-Super-Image2Video when volume and budget do. For dialogue scenes pick Seedance 1.5 Pro, which scores 0 on lipsync accuracy against 0.

### Cosmos3-Super-Image2Video vs LTX-2.3 Pro

- Page: https://vidbenchmark.com/vs/cosmos3-vs-ltx-2-3-pro
- Winner: Cosmos3-Super-Image2Video (Elo 1070)
- Runner-up: LTX-2.3 Pro (Elo 1058)
- Cheaper: Cosmos3-Super-Image2Video at $0.05/second, 38% below LTX-2.3 Pro

Cosmos3-Super-Image2Video wins the head-to-head on blind Elo (1070 vs 1058) and is the safer default for open + strong image2video. Cosmos3-Super-Image2Video is the cheaper option at $0.05 per second, which saves $1.80 per finished minute. Choose Cosmos3-Super-Image2Video when render quality decides the job, and Cosmos3-Super-Image2Video when volume and budget do. For dialogue scenes pick LTX-2.3 Pro, which scores 60 on lipsync accuracy against 0.

### LTX-2.3 Pro vs LTX-2.3 Fast

- Page: https://vidbenchmark.com/vs/ltx-2-3-pro-vs-ltx-2-3-fast
- Winner: LTX-2.3 Pro (Elo 1058)
- Runner-up: LTX-2.3 Fast (Elo 1047)
- Cheaper: LTX-2.3 Fast at $0.04/second, 50% below LTX-2.3 Pro

LTX-2.3 Pro wins the head-to-head on blind Elo (1058 vs 1047) and is the safer default for open-weight quality. LTX-2.3 Fast is the cheaper option at $0.04 per second, which saves $2.40 per finished minute. Choose LTX-2.3 Pro when render quality decides the job, and LTX-2.3 Fast when volume and budget do. For dialogue scenes pick LTX-2.3 Pro, which scores 60 on lipsync accuracy against 59.

### LTX-2.3 Fast vs PixVerse V5.6

- Page: https://vidbenchmark.com/vs/ltx-2-3-fast-vs-pixverse-v5-6
- Winner: LTX-2.3 Fast (Elo 1047)
- Runner-up: PixVerse V5.6 (Elo 1036)
- Cheaper: LTX-2.3 Fast at $0.04/second, 65% below PixVerse V5.6

LTX-2.3 Fast wins the head-to-head on blind Elo (1047 vs 1036) and is the safer default for cheapest usable. LTX-2.3 Fast is the cheaper option at $0.04 per second, which saves $4.50 per finished minute. Choose LTX-2.3 Fast when render quality decides the job, and LTX-2.3 Fast when volume and budget do. For dialogue scenes pick PixVerse V5.6, which scores 61 on lipsync accuracy against 59.

### PixVerse V5.6 vs Hailuo 2.3 Pro

- Page: https://vidbenchmark.com/vs/pixverse-v5-6-vs-hailuo-2-3-pro
- Winner: PixVerse V5.6 (Elo 1036)
- Runner-up: Hailuo 2.3 Pro (Elo 1025)
- Cheaper: Hailuo 2.3 Pro at $0.1/second, 13% below PixVerse V5.6

PixVerse V5.6 wins the head-to-head on blind Elo (1036 vs 1025) and is the safer default for pixverse prior-gen. Hailuo 2.3 Pro is the cheaper option at $0.1 per second, which saves $0.90 per finished minute. Choose PixVerse V5.6 when render quality decides the job, and Hailuo 2.3 Pro when volume and budget do. For dialogue scenes pick PixVerse V5.6, which scores 61 on lipsync accuracy against 0.

### Hailuo 2.3 Pro vs LTX-2 Fast

- Page: https://vidbenchmark.com/vs/hailuo-2-3-pro-vs-ltx-2-fast
- Winner: Hailuo 2.3 Pro (Elo 1025)
- Runner-up: LTX-2 Fast (Elo 1010)
- Cheaper: LTX-2 Fast at $0.04/second, 60% below Hailuo 2.3 Pro

Hailuo 2.3 Pro wins the head-to-head on blind Elo (1025 vs 1010) and is the safer default for minimax classic. LTX-2 Fast is the cheaper option at $0.04 per second, which saves $3.60 per finished minute. Choose Hailuo 2.3 Pro when render quality decides the job, and LTX-2 Fast when volume and budget do. For dialogue scenes pick Hailuo 2.3 Pro, which scores 0 on lipsync accuracy against 0.

### LTX-2 Fast vs Agnes-Video-V2

- Page: https://vidbenchmark.com/vs/ltx-2-fast-vs-agnes-video-v2
- Winner: LTX-2 Fast (Elo 1010)
- Runner-up: Agnes-Video-V2 (Elo 995)
- Cheaper: Agnes-Video-V2 at $0.005/second, 88% below LTX-2 Fast

LTX-2 Fast wins the head-to-head on blind Elo (1010 vs 995) and is the safer default for speed-optimized open. Agnes-Video-V2 is the cheaper option at $0.005 per second, which saves $2.10 per finished minute. Choose LTX-2 Fast when render quality decides the job, and Agnes-Video-V2 when volume and budget do. For dialogue scenes pick LTX-2 Fast, which scores 0 on lipsync accuracy against 0.

### Agnes-Video-V2 vs Wan 2.6

- Page: https://vidbenchmark.com/vs/agnes-video-v2-vs-wan-2-6
- Winner: Agnes-Video-V2 (Elo 995)
- Runner-up: Wan 2.6 (Elo 978)
- Cheaper: Agnes-Video-V2 at $0.005/second, 97% below Wan 2.6

Agnes-Video-V2 wins the head-to-head on blind Elo (995 vs 978) and is the safer default for rock-bottom price. Agnes-Video-V2 is the cheaper option at $0.005 per second, which saves $8.70 per finished minute. Choose Agnes-Video-V2 when render quality decides the job, and Agnes-Video-V2 when volume and budget do. For dialogue scenes pick Wan 2.6, which scores 54 on lipsync accuracy against 0.

### Wan 2.6 vs LTX-2 Pro

- Page: https://vidbenchmark.com/vs/wan-2-6-vs-ltx-2-pro
- Winner: Wan 2.6 (Elo 978)
- Runner-up: LTX-2 Pro (Elo 962)
- Cheaper: LTX-2 Pro at $0.06/second, 60% below Wan 2.6

Wan 2.6 wins the head-to-head on blind Elo (978 vs 962) and is the safer default for alibaba prior-gen. LTX-2 Pro is the cheaper option at $0.06 per second, which saves $5.40 per finished minute. Choose Wan 2.6 when render quality decides the job, and LTX-2 Pro when volume and budget do. For dialogue scenes pick Wan 2.6, which scores 54 on lipsync accuracy against 0.

## Deprecation watch

Sora 2 is scheduled for sunset in September 2026. Pipelines built on it stop returning renders after that date.