Qwen3.5 397B A17B
Qwen · Open weights · Released Feb 16, 2026 · qwen/qwen3.5-397b-a17b
Best $0.39 in / $2.34 out via Alibaba
262K contextVisionToolsReasoningStructured outputPrompt cachingOpen weights
Available from 10 providersUSD / MTOK · PRICES 23 MINUTES AGO · SPEED 2 DAYS AGO
EST. MONTHLY COSTclick a row for detail · click headers to sort
| PROVIDER | INPUT | OUTPUT | CACHE | CONTEXT | QUANT | TTFT | TPS | SMOKE | UPTIME |
|---|---|---|---|---|---|---|---|---|---|
 Alibaba | $0.39 | $2.34 | — | 262K | — | 1.6s | 111 tok/s | 5/5 | 99.8% |
 DeepInfra | $0.45 | $3.00 | $0.22 | 262K | fp8 | ✕ | ✕ | 3/5 | 99.9% |
 Parasail | $0.50 | $3.60 | $0.30 | 262K | fp8 | ✕ | ✕ | 3/5 | 100.0% |
 AtlasCloud | $0.55 | $3.50 | $0.55 | 262K | fp8 | 1.4s | 124 tok/s | 5/5 | 99.2% |
 DigitalOcean | $0.55 | $3.50 | $0.11 | 131K | — | 1.7s | 6 tok/s | 3/5 | 82.1% |
 Phala | $0.55 | $3.50 | $0.225 | 262K | — | 3.2s | 85 tok/s | 3/5 | 100.0% |
 GMICloud | $0.60 | $3.60 | — | 262K | fp8 | 2.2s | 116 tok/s | — | 74.5% |
 Novita | $0.60 | $3.60 | — | 262K | — | 1.3s | 126 tok/s | — | 99.0% |
 StreamLake | $0.60 | $3.60 | $0.12 | 256K | — | 1s | 115 tok/s | — | 99.6% |
 Venice | $0.75 | $4.50 | — | 128K | — | ✕ | ✕ | — | 99.9% |
TTFT/TPS: median (P50) over trailing 504h · chat10k profile (~10K in / 200 out) · measured via OpenRouter routing · ✕ = probed, endpoint returned no measurable stream · SMOKE: of 5 programmatic capability probes passed (JSON schema, instruction following, tool call, long-context, benign compliance; retry-once) · screened via OpenRouter routing · v0.2 · methodology v1.1 · measured 2 days ago
SERVE THIS MODEL AND NOT LISTED? LIST YOUR INFERENCE →
About
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. It delivers...
Specifications
Context window262,144
Max output81,920
Modalities intext, image, video
Modalities outtext
LicenseOpen weights
ReleasedFeb 16, 2026
TokenizerQwen3
WeightsHugging Face ↗
Open weights
safetensors
Related
Embed badge
[](https://modelindex.ai/models/qwen/qwen3.5-397b-a17b)Sources
OpenRouter API23 minutes ago
AGGREGATORHugging Face23 minutes ago
COMMUNITYModelIndex benchmarks8 days ago
MEASUREDVerified 23 minutes ago · list prices, not negotiated rates.