Skip to main content
  • Pricing

Oakgen Arena / Model intelligence

Find your next
creative advantage.

Compare AI image and video models.
Start with the rankings. Choose for your craft.

September 2026 benchmark snapshot

A starting point for your next project

See where models stand.

12 task boards · Source index v1.0

video / General capabilities

Video models, compared.

Swipe to compare all models →

Oakgen Arena · v1.0
Seedance 2.5
Wan 3.0
Seedance 2.0
Seedance 2.0 Mini
Google Omni Flash
Flux 3 Video
MiniMax H3Anchor
Kling 3.0 Omni
HappyHorse 1.1
Grok Imagine 1.5
↥ 95% confidence interval · Anchor interval not reported.10 ranked models

Relative scores on a zero-based display scale. Compare within this board only. Values are rounded for display.

Top 10 of 11 models · Relative scoresFull leaderboard

September 2026 benchmark snapshot · Curated by Oakgen. Data source and limits.

image / General capabilities

Image models, compared.

Swipe to compare all models →

Oakgen Arena · v1.0
Seedream 5.0 Pro
GPT Image 2
Nano Banana Pro
Grok Imagine 2.0Anchor
Nano Banana 2
Qwen Image 3.0
Flux.2 Pro
↥ 95% confidence interval · Anchor interval not reported.7 ranked models

Relative scores on a zero-based display scale. Compare within this board only. Values are rounded for display.

Top 7 of 7 models · Relative scoresFull leaderboard

September 2026 benchmark snapshot · Curated by Oakgen. Data source and limits.

A score needs context.

Read the evidence.
Then make it your own.

These are dated, third-party reference rankings. Oakgen adds guidance to help you read them and test what matters to your work.

Inside the methodology
01

Choose the right task

Compare within a board. Ads, film and image editing ask different things of a model.

02

Look beyond the rank

Read uncertainty alongside scores. Small gaps do not settle every creative decision.

03

Try your own brief

Test your prompts, settings and assets before committing to a production workflow.

Your idea is the real test.

Turn a shortlist into a first draft.

Compare your own prompt