qwen/qwen3.5-397b-a17bQwen3.5 397B-A17B is Alibaba’s open-weights AI model, released Feb 13, 2026. It costs $0.39/M input and $2.34/M output tokens, and handles a 262K-token context window. Its strongest use case is agentic & tools, where it ranks in the 87th percentile of the models Model Beat tracks.
The Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency.
Scores on standardized evaluations. Higher is better — and rank shows where Qwen3.5 397B-A17B lands among all models tracked on Model Beat.
Graduate-level scientific reasoning
Frontier of human expert knowledge
Scientific research coding
Tool-agent-user reliability
The Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency.
Qwen3.5 397B-A17B is an AI model developed by Alibaba, part of the Qwen family, released Feb 13, 2026.
Qwen3.5 397B-A17B's list price from Alibaba is $0.39 per million input tokens and $2.34 per million output tokens. The cheapest credible third-party provider on OpenRouter (DigitalOcean) serves it at $0.39/$2.45 per 1M.
Qwen3.5 397B-A17B supports a context window of up to 262K tokens.
On standardized evaluations tracked by Epoch AI, Qwen3.5 397B-A17B scores 95.6% on τ²-bench, 89.3% on GPQA Diamond, 27.3% on Humanity's Last Exam.
Qwen3.5 397B-A17B is released with open weights (Open weights (unrestricted)), so the model can be downloaded and self-hosted.
Benchmark data from Epoch AI (CC BY) and Artificial Analysis; pricing, specs & descriptions from OpenRouter. Data updated Jul 30, 2026.