As of 5 November 2026, no option is proven ahead. 4 options of 16 cannot be separated with the evidence in hand: gemini-omni-1.1-flash, gemini-omni-flash, flux-3-video and grok-imagine-video-1.5-agent.
- Options
- 16
- Measured
- 16
- Shortlist
- 4
- Plotted
- 3
- Priced
- 3
The job
Generate short video from a text prompt through a vendor's own API, and judge the model by how often people prefer its clips in blind comparisons on Arena, against what a minute of generated video costs.
Ranking
16 options measured, shortlisted rows shaded
| # | Option | Score 0 to 100 | Cost per minute of video | Interval | Price page |
|---|---|---|---|---|---|
| 1 | gemini-omni-1.1-flashgoogleFrontierShortlist | 91.6 | $6.08 | ci9587.7 to 95.4 | |
| 2 | gemini-omni-flashgoogleShortlist | 90.7 | Not priced | ci9588.5 to 93.0 | |
| 3 | flux-3-videobflShortlist | 85.8 | Not priced | ci9581.5 to 90.1 | |
| 4 | grok-imagine-video-1.5-agentxaiShortlist | 85.6 | Not priced | ci9581.0 to 90.2 | |
| 5 | dreamina-seedance-2.0-720pbytedance | 82.3 | Not priced | ci9580.3 to 84.3 | |
| 6 | wan3.0alibaba | 81.6 | Not priced | ci9578.4 to 84.7 | |
| 7 | dreamina-seedance-2.5-720pbytedance | 81.0 | Not priced | ci9578.8 to 83.1 | |
| 8 | minimax-h3minimaxFrontier | 77.6 | $4.80 | ci9575.4 to 79.7 | |
| 9 | muse-videometa | 76.5 | Not priced | ci9572.8 to 80.1 | |
| 10 | happyhorse-1.0aorizon | 69.3 | Not priced | ci9566.2 to 72.4 | |
| 11 | sora-2-proopenai | 54.4 | Not priced | ci9552.8 to 56.1 | |
| 12 | veo-3.1-audiogoogle | 53.5 | Not priced | ci9549.9 to 57.1 | |
| 13 | veo-3.1-audio-1080pgoogle | 53.1 | Not priced | ci9550.5 to 55.6 | |
| 14 | veo-3.1-fast-audiogoogle | 53.0 | Not priced | ci9550.4 to 55.6 | |
| 15 | veo-3.1-fast-audio-1080pgoogle | 52.0 | Not priced | ci9549.5 to 54.5 | |
| 16 | grok-imagine-video-720pxaiFrontier | 48.0 | $3.00 | ci9546.4 to 49.6 |
Order is by score, highest first. Intervals overlap, so neighbouring ranks do not differ with certainty. Cost is in USD per minute of video; bars start at zero. Interval: ci95, built from 95% confidence intervals given by the source.
Price against performance
Score against cost of one minute of video, log scale
- Frontier: no other option is both cheaper and scores higher
- Other option
- Ring: on the shortlist, not separable with certainty
- Line: score interval, ci95 (built from 95% confidence intervals given by the source)
13 options not plotted: a coordinate is unknown
- dreamina-seedance-2.0-720p: Cost unknown
- dreamina-seedance-2.5-720p: Cost unknown
- flux-3-video: Cost unknown
- gemini-omni-flash: Cost unknown
- grok-imagine-video-1.5-agent: Cost unknown
- happyhorse-1.0: Cost unknown
- muse-video: Cost unknown
- sora-2-pro: Cost unknown
- veo-3.1-audio: Cost unknown
- veo-3.1-audio-1080p: Cost unknown
- veo-3.1-fast-audio: Cost unknown
- veo-3.1-fast-audio-1080p: Cost unknown
- wan3.0: Cost unknown
Cost for this job
Date read and source for all 16 options
Cost of one minute of video, in USD per minute of video, cheapest first within each group.
Workload and assumptions
Cost of one minute (60 seconds) of generated video at about 720p (768P for MiniMax), as the vendor's per-second price times 60. For Gemini Omni the vendor prices video output tokens, so the price is the page's 5,792 tokens per second of 720p video times its price per token. Audio, upscaling, extensions and failed or discarded clips are not counted. OpenAI's page says Sora 2 was shut down on 24 September 2026 and Google's page says the Veo 3.1 previews were shut down on 22 October 2026, so those models have unknown cost.
| Option | Cost | Price read | Source | Notes |
|---|---|---|---|---|
| Cost 3 | ||||
grok-imagine-video-720pxai | $3.00 | 0 days old | ||
minimax-h3minimax | $4.80 | 0 days old | ||
gemini-omni-1.1-flashgoogle | $6.08 | 0 days old | ||
| No cost available 13 | ||||
dreamina-seedance-2.0-720pbytedance | Unknown | Not read | No source | 1 note
|
dreamina-seedance-2.5-720pbytedance | Unknown | Not read | No source | 1 note
|
flux-3-videobfl | Unknown | Not read | No source | 1 note
|
gemini-omni-flashgoogle | Unknown | Not read | No source | 1 note
|
grok-imagine-video-1.5-agentxai | Unknown | Not read | No source | 1 note
|
happyhorse-1.0aorizon | Unknown | Not read | No source | 1 note
|
muse-videometa | Unknown | Not read | No source | 1 note
|
sora-2-proopenai | Unknown | Not read | No source | 1 note
|
veo-3.1-audiogoogle | Unknown | Not read | No source | 1 note
|
veo-3.1-audio-1080pgoogle | Unknown | Not read | No source | 1 note
|
veo-3.1-fast-audiogoogle | Unknown | Not read | No source | 1 note
|
veo-3.1-fast-audio-1080pgoogle | Unknown | Not read | No source | 1 note
|
wan3.0alibaba | Unknown | Not read | No source | 1 note
|
What evidence would settle it
The three conditions and what is missing
- flux-3-video: another independent origin or repeats (interval 8.6 points wide), metric Arena text-to-video rating (Elo-style, overall category).
- gemini-omni-1.1-flash: another independent origin or repeats (interval 7.7 points wide), metric Arena text-to-video rating (Elo-style, overall category).
- gemini-omni-flash: another independent origin or repeats (interval 4.5 points wide), metric Arena text-to-video rating (Elo-style, overall category).
- grok-imagine-video-1.5-agent: another independent origin or repeats (interval 9.1 points wide), metric Arena text-to-video rating (Elo-style, overall category).
The three conditions for naming a best option
- The leader's lower bound must exceed every other option's upper bound by at least 3 points. Not met: the closest gap is -5.3 points (gemini-omni-1.1-flash against gemini-omni-flash).
- Every critical metric must come from two independent origins, or from one auditable test of our own. Not met: 1 origin and no auditable own test for gemini-omni-1.1-flash.
- The price must be verified within 30 days. Met: read 0 days before this date.
Sources and limits
4 sources, the method notes and this board as JSON
Scores on this board come from Arena Leaderboard Dataset (lmarena-ai/leaderboard-dataset). We rescale them to a 0 to 100 scale against fixed anchors (Arena text-to-video rating (Elo-style, overall category): 1150 counts as 0 and 1550 counts as 100), so a new option never moves another option's score. The licence and our changes are stated in the data credit at the foot of this page.
- Arena Leaderboard Dataset (lmarena-ai/leaderboard-dataset) free to use with credit; checked
- Gemini API pricing prices only, as facts with a link; checked
- MiniMax video pay-as-you-go pricing prices only, as facts with a link; checked
- xAI API models and pricing prices only, as facts with a link; checked
Notes on this board. Only one origin (Arena) measures quality, so no board entry can be named Best until a second independent origin or an own test exists. Arena's rating comes from blind votes by its own users on their own prompts; it is not a test of your job. Prices are read from each vendor's own API price page and are list prices for the stated workload; a model whose price could not be read on the vendor's page is listed with unknown cost and is not plotted. Selection: the 15 models with the highest Arena rank in the overall category, plus every model ranked 16 to 48 whose price was read on a vendor page and is at most $4 per minute of video.
Scores are on a 0 to 100 scale against editorial anchors that are not yet validated. Prices are for the workload above and change without notice. Evidence older than 60 days is left out, and prices older than 30 days are flagged.
This board as JSON, with every input id, anchor and weight.