qwen/qwen3.6-27b
Qwen3.6 27B is a dense 27-billion-parameter language model from the Qwen Team at Alibaba, released in April 2026. It features hybrid multimodal capabilities — accepting text, image, and video inputs...
| Provider | In $/M | Out $/M | Context | Uptime |
|---|---|---|---|---|
| Io Netfp8 | $0.288 | $3.17 | 262K | 100% |
| Morph | $0.289 | $2.40 | 131K | 97% |
| Chutesfp8 | $0.300 | $2.00 | 262K | 99.9% |
| SiliconFlowfp8 | $0.300 | $3.20 | 262K | — |
| Phala | $0.320 | $2.70 | 262K | — |
| DeepInfrafp8 | $0.320 | $3.20 | 262K | 99.9% |
| Venicefp8 | $0.325 | $3.25 | 256K | — |
| Alibaba | $0.450 | $2.70 | 262K | 96.8% |
| WandBfp8 | $0.600 | $3.60 | 262K | — |
Qwen3.6 27B costs $0.288 per million input tokens and $3.17 per million output tokens via OpenRouter, making it 125th cheapest of 298 paid models.
Qwen3.6 27B scores 37.1 on the Artificial Analysis Intelligence Index, ranking 55th of 178 benchmarked models, with a GPQA Diamond score of 83%.
Qwen3.6 27B generates around 80 tokens per second with 390ms time-to-first-token (p50), the 92nd fastest tracked model.
Qwen3.6 27B supports a 262K-token context window and can output up to 262K tokens. It accepts text, image, video input.