Alibaba · qwen
Qwen3.5 9B
Qwen instruction model for multilingual chat, reasoning, and tool use
$0.10 in · $0.15 out
per 1M tokens · provider rate closest to the median. No first-party Alibaba price was listed. The reference price is Deep Infra's rate, the provider closest to the median.
Provider input prices run from $0.05 to $0.30 per 1M tokens. Output prices run from $0.13 to $0.48.
- Context
- 262,144
- Max output
- 65,536
- Cache read
- —
- Cache write
- —
- Reasoning
- Yes
- Tool calls
- Yes
- Structured output
- Yes
- Vision
- Yes
- Open weights
- Yes
- License
- —
- Released
- Feb 23, 2026
- Updated in source
- Feb 23, 2026
- Knowledge
- —
- Providers
- 15
Providers
| Provider | Input | Output | Cache read | Cache write | Status |
|---|---|---|---|---|---|
| NanoGPT | $0.05 | $0.15 | $0.025 | — | — |
| EmpirioLabs AI | $0.09 | $0.13 | $0.045 | — | — |
| Merge Gateway | $0.09 | $0.13 | — | — | — |
| Deep Infrareference | $0.10 | $0.15 | — | — | — |
| DevPass | $0.10 | $0.15 | — | — | — |
| Kilo Gateway | $0.10 | $0.15 | — | — | — |
| LLM Gateway | $0.10 | $0.15 | — | — | — |
| OpenRouter | $0.10 | $0.15 | — | — | — |
| Venice AI | $0.10 | $0.15 | — | — | — |
| Cortecs | $0.11 | $0.17 | — | — | — |
| TensorX | $0.15 | $0.20 | $0.038 | $0.19 | — |
| routing.run | $0.16 | $0.48 | — | — | — |
| Hugging Face | $0.17 | $0.25 | — | — | — |
| Pioneer | $0.30 | $0.30 | $0.30 | $0.30 | — |
| QVAC | $0 | $0 | — | — | — |
Jobs this model qualifies for
- Coding agents — clears the bar
- Long-document Q&A — clears the bar
- Cheapest summarization — clears the bar