qwen/qwen3.5-9b
Qwen3.5-9B is a multimodal foundation model from the Qwen3.5 family, designed to deliver strong reasoning, coding, and visual understanding in an efficient 9B-parameter architecture. It uses a unified vision-language design...
| Provider | In $/M | Out $/M | Context | Uptime |
|---|---|---|---|---|
| SiliconFlowfp8 | $0.100 | $0.150 | 262K | 98.3% |
| DeepInfrabf16 | $0.100 | $0.150 | 262K | 99.7% |
| Venicefp8 | $0.100 | $0.150 | 256K | 99.8% |
| Together | $0.170 | $0.250 | 262K | 99.6% |
Qwen3.5-9B costs $0.100 per million input tokens and $0.150 per million output tokens via OpenRouter, making it 49th cheapest of 298 paid models.
Qwen3.5-9B scores 32.4 on the Artificial Analysis Intelligence Index, ranking 68th of 180 benchmarked models, with a GPQA Diamond score of 81%.
Qwen3.5-9B generates around 88 tokens per second with 568ms time-to-first-token (p50), the 78th fastest tracked model.
Qwen3.5-9B supports a 262K-token context window and can output up to 262K tokens. It accepts text, image, video input.