Qwen3.6 27B
by Abacus · family qwen · listed Apr 2026RTW
Qwen vision-language model for visual reasoning, documents, and agent tasks
Current rates USD per 1M tokens
Input$0.32
Output$3.20
Cache read—
Cache write—
Against the market
On the record
- Model id
Qwen/Qwen3.6-27B- Context window
- 262K tokens
- Max output
- 8K tokens
- Input modalities
- text
- Output modalities
- text
- Reasoning
- yes
- Tool calls
- yes
- Open weights
- yes
- Knowledge cutoff
- —
- Released
- 2026-04-22
- Last updated
- 2026-04-22
- Provider docs
- abacus.ai/help/api
Back of the envelope
≈
Also listed at
- CrofAI$0.20 in · $1.50 out
- NanoGPT$0.203 in · $2.24 out
- OpenRouter$0.285 in · $2.40 out
- Vultr$0.30 in · $2.00 out
- SiliconFlow$0.30 in · $3.20 out
- Deep Infra$0.32 in · $3.20 out
- Kilo Gateway$0.325 in · $3.25 out
- Synthetic$0.45 in · $3.60 out
- Hugging Face$0.47 in · $3.19 out
- OVHcloud AI Endpoints$0.47 in · $3.19 out
- STACKIT$0.53 in · $0.76 out
- Vercel AI Gateway$0.60 in · $3.60 out
- Pioneer$0.60 in · $0.60 out
- Weights & Biases$0.60 in · $3.60 out
- Alibaba$0.60 in · $3.60 out