qwen/qwen3.8-27b

Qwen 3.8 27B via OpenRouter — a fast dense model with tool calling and a tunable reasoning effort; pinned at order -3 so it is the deployment default ahead of z-ai/glm-5.3 (-2), whose always-on max-effort reasoning made rounds take minutes.

• Provider: OpenRouter
• Endpoint: https://openrouter.ai/api/v1
• Provider ref: Provider/OpenRouter
• Order: -3
• Input price per million tokens: 0.42
• Output price per million tokens: 3.0
• Currency: USD
• Capability: 0
• Supports tools: yes