Qwen: Qwen 3.5 Plus
by Kilo Gateway · family qwen · listed Feb 2026RTV
The Qwen3.5 native vision-language series Plus models are built on a hybrid architecture that integrates linear attention mechanisms with sparse mixture-of-experts models, achieving higher inference efficiency. In a variety of task evaluations, the 3.5 series consistently demonstrates performance on par with state-of-the-art leading models. Compared to the 3 series, these models show a leap forward in both pure-text and multimodal capabilities.
Current rates USD per 1M tokens
Input$0.40
Output$2.40
Cache read$0.04
Cache write$0.50
Against the market
On the record
- Model id
qwen/qwen3.5-plus- Context window
- 1M tokens
- Max output
- 64K tokens
- Input modalities
- text, image, pdf
- Output modalities
- text
- Reasoning
- yes
- Tool calls
- yes
- Open weights
- no
- Knowledge cutoff
- Apr 2025
- Released
- 2026-02-16
- Last updated
- 2026-02-16
- Provider docs
- kilo.ai
Back of the envelope
≈
Also listed at
- Alibaba Coding Planfree in · free out
- Alibaba Coding Plan (China)free in · free out
- AIHubMix$0.11 in · $0.658 out
- Alibaba (China)$0.115 in · $0.688 out
- Merge Gateway$0.115 in · $0.688 out
- 302.AI$0.12 in · $0.69 out
- OpenCode$0.20 in · $1.20 out
- OrcaRouter$0.40 in · $2.40 out
- Alibaba$0.40 in · $2.40 out
- Vercel AI Gateway$0.40 in · $2.40 out
- LLMTR$0.40 in · $2.40 out
- ZenMux$0.40 in · $2.40 out
- Meganova$0.40 in · $2.40 out
- Ofox$0.40 in · $2.40 out
- NanoGPT$0.40 in · $2.40 out