#1
Claude Fable 5
Anthropic
Leads the current blind-preference vision leaderboard.
Arena score1312±9
- visual prompt reasoning
- strong vision preference
The model that wins your task, wired into your business.
No single model wins everything. Today the leader in code is not the leader in voice, the best model for documents loses on images, and half of these positions move within a month. That is what the board below is for: the top 3 for each of 26 tasks, rebuilt every 10 days from public leaderboards and independent tests.
Then I build the thing that uses it. Agents that handle support, pipelines that read invoices and contracts, voice that answers the phone, bots that qualify leads while you sleep. It plugs into what you already run — CRM, spreadsheets, Telegram, your own API — and every step calls the model that wins that particular step, whichever provider it belongs to.
Write me on Telegram — @simvimTell me which process eats your week. I answer myself.
Image understanding
For reasoning about photos, diagrams and visual content.
Second and third are practically tied within the reported intervals.
#2
Qwen3.8 Max
Alibaba
Places narrowly ahead of third on mixed visual prompts.
Arena score1302±8
- mixed image reasoning
- competitive vision score
#3
Claude Opus 4.7
Anthropic
Scores effectively level with second place on visual prompts.
Arena score1301±7
- visual reasoning quality
- strong multimodal preference
Text from image (OCR)
For reading text embedded in images and scanned content.
OCR Arena is preference-based rather than a character-error-rate benchmark.
#1
Claude Fable 5
Anthropic
Leads the current OCR-specific Arena preference slice.
Arena score1329±9
- image text reading
- strong ocr preference
#2
Qwen3.8 Max
Alibaba
Places second on OCR-oriented visual prompts.
Arena score1317±9
- visual text extraction
- competitive ocr preference
#3
Claude Opus 4.7
Anthropic
Completes the current leading OCR preference cluster.
Arena score1314±7
- image text reasoning
- stable ocr preference
Image generation
For generating images from natural-language prompts.
Second and third differ by one displayed Elo point.
#1
GPT Image 2
OpenAI
Leads the selected live text-to-image preference leaderboard.
Elo1369
- text image preference
- current elo lead
#2
Reve 2.1
Reve
Places second on the selected live image arena.
Elo1321
- prompt image quality
- strong arena preference
#3Google
Nano Banana 2 (Gemini 3.1 Flash Image Preview)
Preview
Scores one Elo point behind second on the selected live board.
Elo1320
- prompt image generation
- competitive arena score
Image editing
For editing existing images from text instructions.
The two major public image-edit arenas currently disagree materially.
#1
GPT Image 2
OpenAI
Leads the image-editing preference board.
Arena score1463±4
- instruction based editing
- fresh arena lead
#2
Grok Imagine Image 2.0
xAI
Preliminary
Places second on the selected fresh image-edit arena.
Arena score1439±8
- image edit preference
- strong arena placement
#3
MAI-Image-2.6 Preview
Microsoft AI
Preview
Ranks third on the same image-editing board.
Arena score1420±8
- instruction image editing
- competitive arena score
How the ranking is built
There is no single “best AI model” here. Every task is judged on its own: public leaderboards, independent tests and official model data, weighed together rather than copied from one board. The whole thing is re-checked every 10 days. Two modes of the same base model never take two places in one top 3. Likes are a separate reader signal — they never move the ranking.