QQwen | Qwen3 Coder
Qwen3-Coder-480B-A35B-Instruct is a Mixture-of-Experts (MoE) code generation model developed by the Qwen team. It is optimized for agentic coding tasks such as function calling, tool use, and long-context reasoning over repositories. The model features 480 billion total parameters, with 35 billion active per forward pass (8 out of 160 experts). Pricing for the Alibaba endpoints varies by context length. Once a request is greater than 128k input tokens, the higher pricing is used.
Fair
Significant guardrail issues may impede work.
Strengths
No notable strengths identified
Weaknesses
- Weak doctrinal alignment
- Resists worldview confession
- Struggles with 1.1. Missiological Research
- Struggles with 1.3. Apologetics
- Struggles with 2.5. Call to Repentance
Qwen3-Coder-480B-A35B-Instruct is a Mixture-of-Experts (MoE) code generation model developed by the Qwen team. It is optimized for agentic coding tasks such as function calling, tool use, and long-context reasoning over repositories. The model features 480 billion total parameters, with 35 billion active per forward pass (8 out of 160 experts). Pricing for the Alibaba endpoints varies by context length. Once a request is greater than 128k input tokens, the higher pricing is used.
Qwen
qwen/qwen3-coder
1
Categories
Recent Tests
| Date | Score | Version | Tier 1 (Task) | Tier 2 (Gospel) | Tier 3 (Worldview) | Trust Tier |
|---|---|---|---|---|---|---|
| 10/11/2026 | 52.7 | 1.0.0 | 55.2 | 46.7 | 46.7 | automated |