Qwen3.6 Max Preview
Compared with the previously released Qwen3-Max and Qwen3.6-Plus, this model features enhanced vibe coding abilities, more efficient coding agent execution, and significantly improved front-end development skills.
text → text · made by Alibaba
$1.03→$6.16per Mtok in / out
Nous Research
Sold by 3 ways
| Seller | Lane | Rate |
|---|---|---|
| Nous Research 0 ways | $1.03per Mtok in$6.16per Mtok out$1.28per Mtok cache write | |
| Nous Research aggregatornot on the seller's list today | standard | $1.03per Mtok in$6.16per Mtok out$1.28per Mtok cache writeinference-api.nousresearch.com · read 2026-09-12 |
| OpenRouter 0 ways | $1.03per Mtok in$6.16per Mtok out$1.28per Mtok cache write | |
| OpenRouter aggregatornot on the seller's list today | standard | $1.03per Mtok in$6.16per Mtok out$1.28per Mtok cache writeopenrouter.ai · read 2026-08-24 |
| Vercel AI Gateway 0 ways | $1.3per Mtok in$7.8per Mtok out$0.13per Mtok cached$1.62per Mtok cache write | |
| Vercel AI Gateway aggregatornot on the seller's list today | standard | $1.3per Mtok in$7.8per Mtok out$0.13per Mtok cached$1.62per Mtok cache writeai-gateway.vercel.sh · read 2026-08-25 |
Measured 13 standings
| Place | Board | Metric | Score |
|---|---|---|---|
| 9thof 35 | Swe bench verified — Epoch AI | mean_score (max) | 76.653 |
| 20thof 102 | Simplebench — Epoch AI | Score (AVG@5) (max) | 0.63 |
| 25thof 93 | Simpleqa verified — Epoch AI | mean_score (max) | 52.004 |
| 35thof 101 | FrontierMath (Tiers 1-3) | mean_score (max) | 23.103 |
| 35thof 72 | Frontiermath tier 4 — Epoch AI | mean_score (max) | 4.167 |
| 42ndof 71 | Vending bench 2 — Epoch AI | Score (max) | 4254.187 |
| 59thof 414 | LMArena · Text | Score (Elo) (max) | 1460 |
| 62ndof 141 | LMArena · WebDev | Score (Elo) (max) | 1482 |
| 62ndof 135 | Webdev arena — Epoch AI | Arena Score (max) | 1479.06 |
| 75thof 306 | OTIS Mock AIME 2024-2025 | mean_score (max) | 91.111 |
| 78thof 326 | GPQA Diamond | mean_score (max) | 87.374 |
| 101stof 197 | Lmca — Epoch AI | Score (max) | 42.53 |
| 102ndof 235 | Dtbench — Epoch AI | Accuracy (max) | 87.2 |
About
Qwen3.6 Max Preview — a text model from Alibaba, placed 59th of 414 on LMArena · Text.
It takes text and returns text, with a context window of 240,000 tokens. It was published in April 2026, trained on material up to April 2025. Its sellers say it can reason step by step and call a tool. The catalogue files it under reasoning, code and agents. It has been measured on 13 boards, and stands best at 59th of 414 on LMArena · Text.
Every current figure
- Maker
- Alibaba
- Register
- model
- Takes
- text
- Returns
- text
- Context
- 240,000 tokens
- Longest answer
- 65,536 tokens
- Published
- April 2026
- Knowledge to
- April 2025
- Licence
- not read
- Boards
- 13
- Best place
- 59th of 414 — LMArena · Text
Known as 3 names
Qwen 3.6 Max Previewqwen/qwen3.6-max-previewalibaba/qwen-3.6-max-preview