Qwen3.5 397B A17B pricing & benchmarks
Released Feb 16, 2026 · served by 11 providers · ranked #47 by intelligence in our index
What Qwen3.5 397B A17B is
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...
- Model ID:
qwen/qwen3.5-397b-a17b - Modality: text+image+video->text
- Tool calling: supported · Structured output: supported
- Weights:
Qwen/Qwen3.5-397B-A17Bon Hugging Face
Providers and prices for Qwen3.5 397B A17B
The same model costs different amounts depending on who serves it. Prices are per 1M tokens, cheapest first. Uptime and throughput are OpenRouter's rolling measurements, not vendor claims.
| Provider | Input | Output | Context | Quant | Uptime 24h | Throughput |
|---|---|---|---|---|---|---|
| Alibabacheapest | $0.390 | $2.34 | 262K | — | 99.9% | — |
| DigitalOcean | $0.385 | $2.45 | 131K | — | 96.6% | — |
| Chutes | $0.450 | $3.00 | 262K | fp8 | 99.5% | — |
| DeepInfra | $0.450 | $3.00 | 262K | fp8 | 98.9% | — |
| AtlasCloud | $0.550 | $3.50 | 262K | fp8 | 99.5% | — |
| Phala | $0.550 | $3.50 | 262K | — | 98.9% | — |
| Parasail | $0.500 | $3.60 | 262K | fp8 | 99.7% | — |
| Novita | $0.600 | $3.60 | 262K | — | 99.8% | — |
| StreamLake | $0.600 | $3.60 | 256K | — | 99.4% | — |
| GMICloud | $0.600 | $3.60 | 262K | fp8 | 0.0% | — |
| Venice | $0.750 | $4.50 | 128K | — | 93.0% | — |
Qwen3.5 397B A17B benchmark results
Design Arena head-to-head results, as reported through the OpenRouter model API.
| Category | Arena | Rank | Elo | Win rate |
|---|---|---|---|---|
| svg | models | #33 | 1191 | 56.1% |
| 3d | models | #41 | 1218 | 56.7% |
| codecategories | models | #43 | 1210 | 52.6% |
| dataviz | models | #45 | 1204 | 53.2% |
| website | models | #45 | 1215 | 52.6% |
| gamedev | models | #51 | 1196 | 50.1% |
| uicomponent | models | #52 | 1199 | 51.4% |
Cheaper models in the same class
Models scoring within 4 points of Qwen3.5 397B A17B on the Intelligence Index, but with a lower output price.
FAQ
How much does the Qwen3.5 397B A17B API cost?
$0.385 per 1M input tokens and $2.34 per 1M output tokens. A typical workload of 1M input + 300K output tokens costs about $1.09.
Is Qwen3.5 397B A17B free?
No. It is a paid model starting at $0.385 per 1M input tokens, though some providers offer trial credits.
Which provider is cheapest for Qwen3.5 397B A17B?
Alibaba at $0.390 input / $2.34 output per 1M tokens.
Can I self-host Qwen3.5 397B A17B?
Yes — weights are published as Qwen/Qwen3.5-397B-A17B on Hugging Face, so you can run it on your own hardware.