Qwen3.6 27B
The Qwen3.6 35B-A3B native vision-language model is built on a hybrid architecture that integrates linear attention mechanisms with a sparse mixture-of-experts framework, achieving higher inference efficiency.
text + image + file → text · made by Alibaba
Free
Offered at no charge, which is not the same as cheap: a free lane carries a rate limit and can be withdrawn. The prices above are what you pay when it is not available to you.
- Groq
Sold by 22 ways
| Seller | Lane | Rate |
|---|---|---|
| Groq api | standard | $0.6per Mtok in$3per Mtok out$0.3per Mtok cachedconsole.groq.com · read 2026-08-25 |
| Thinking Machines Lab api | training | $1.86per Mtok in$5.59per Mtok out$0.37per Mtok cachedtinker-docs.thinkingmachines.ai · read 2026-08-25 |
| Groq aggregator | free | $0per Mtok in$0per Mtok outconsole.groq.com · read 2026-09-16 |
| ElectronHub aggregator | standard | $0.3per Mtok in$2.5per Mtok outapi.electronhub.ai · read 2026-09-16 |
| Nous Research aggregator | standard | $0.3per Mtok in$2per Mtok out$0.03per Mtok cachedinference-api.nousresearch.com · read 2026-09-12 |
| OpenRouter 7 ways | $0.3per Mtok in$2per Mtok out$0.03per Mtok cached | |
| OpenRouter aggregator | standard | $0.3per Mtok in$2per Mtok out$0.03per Mtok cachedopenrouter.ai · read 2026-09-05 |
| OpenRouter aggregator | chutes/fp8 | $0.3per Mtok in$2per Mtok out$0.03per Mtok cachedopenrouter.ai · read 2026-08-25 |
| OpenRouter aggregator | siliconflow/fp8 | $0.3per Mtok in$3.2per Mtok outopenrouter.ai · read 2026-08-25 |
| OpenRouter aggregator | deepinfra/fp8 | $0.32per Mtok in$3.2per Mtok outopenrouter.ai · read 2026-08-29 |
| OpenRouter aggregator | phala | $0.32per Mtok in$2.7per Mtok out$0.15per Mtok cachedopenrouter.ai · read 2026-08-25 |
| OpenRouter aggregator | venice/fp8 | $0.33per Mtok in$3.25per Mtok outopenrouter.ai · read 2026-08-25 |
| OpenRouter aggregator | alibaba | $0.45per Mtok in$2.7per Mtok outopenrouter.ai · read 2026-08-25 |
| OpenRouter aggregatornot on the seller's list today | coreweave/fp8 | $0.6per Mtok in$3.6per Mtok out$0.12per Mtok cachedopenrouter.ai · read 2026-08-25 |
| SiliconFlow aggregator | standard | $0.3per Mtok in$3.2per Mtok outmodels.dev · read 2026-09-16 |
| DeepInfra aggregator | standard | $0.32per Mtok in$3.2per Mtok outapi.deepinfra.com · read 2026-08-25 |
| Hugging Face Inference Providers 3 ways | $0.47per Mtok in$3.19per Mtok out | |
| Hugging Face Inference Providers aggregator | deepinfra | $0.32per Mtok in$3.2per Mtok outrouter.huggingface.co · read 2026-08-25 |
| Hugging Face Inference Providers aggregator | standard | $0.47per Mtok in$3.19per Mtok outmodels.dev · read 2026-09-16 |
| Hugging Face Inference Providers aggregator | ovhcloud | $0.47per Mtok in$3.19per Mtok outrouter.huggingface.co · read 2026-08-25 |
| Venice AI aggregator | standard | $0.33per Mtok in$3.25per Mtok outmodels.dev · read 2026-09-16 |
| Novita AI aggregator | standard | $0.6per Mtok in$3.6per Mtok outapi.novita.ai · read 2026-08-25 |
| Vercel AI Gateway aggregator | standard | $0.6per Mtok in$3.6per Mtok outai-gateway.vercel.sh · read 2026-08-25 |
| PPIO aggregator | standard | $3per Mtok in$18per Mtok outapi.ppinfra.com · read 2026-09-16 |
Measured 14 standings
| Place | Board | Metric | Score |
|---|---|---|---|
| 15thof 17 | SWE-rebench Leaderboard (default window 15/05/2026–01/07/2026, 111 problems) | Resolved Rate (%) | 31.2 |
| 54thof 106 | Frontiermath tiers 1 3 v2 — Epoch AI | mean_score | 35.088 |
| 55thof 123 | Lmca — Epoch AI | Score | 34.53 |
| 60thof 290 | OTIS Mock AIME 2024-2025 | mean_score | 91.111 |
| 72ndof 222 | Chess puzzles — Epoch AI | mean_score | 22 |
| 72ndof 161 | Dtbench — Epoch AI | Accuracy | 78.13 |
| 74thof 143 | Artificial Analysis · Agentic Index | Index | 20.1 |
| 76thof 313 | GPQA Diamond | mean_score | 85.859 |
| 77thof 184 | Artificial Analysis · Coding Index | Index | 53.7 |
| 109thof 302 | Artificial Analysis · Intelligence Index | Artificial Analysis Intelligence Index | 22 |
| 112thof 127 | Mystery game puzzles — Epoch AI | mean_score (none) | 7 |
| 117thof 177 | Critpt — Epoch AI | Accuracy (none) | 0.857 |
| 130thof 168 | Scicode — Epoch AI | Score (none) | 37.269 |
| 287thof 656 | Epoch capabilities index — Epoch AI | ECI Score (none) | 146.46 |
About
Qwen3.6 27B — a text model from Alibaba, sold by 12 companies from $0.3 in and $2.5 out per million tokens, placed 60th of 290 on OTIS Mock AIME 2024-2025.
It takes text, images and files and returns text, with a context window of 256,000 tokens. It was published in April 2026. Its sellers say it can reason step by step and call a tool. The catalogue files it under extract, reasoning and code. Twelve companies sell it. The cheapest is $0.3 in and $2.5 out per million tokens at ElectronHub. Beside the standard rate there are separately routed and regional lanes. It has been measured on 14 boards, and stands best at 60th of 290 on OTIS Mock AIME 2024-2025.
Every current figure
- Maker
- Alibaba
- Register
- model
- Takes
- text + image + file
- Returns
- text
- Context
- 256,000 tokens
- Published
- April 2026
- Parameters
- 28 billion
- Licence
- not read
- Sellers
- 12
- Maker's own price
- not read
- Price
- $0.3 in and $2.5 out per million tokens — ElectronHub
- Boards
- 14
- Best place
- 60th of 290 — OTIS Mock AIME 2024-2025
Known as 3 names
Qwen3.6-27BQwen/Qwen3.6-27Balibaba/qwen3.6-27b