ModelRefs / Qwen 3 Coder by Alibaba — Benchmarks, Pricing & Review (202…

Qwen 3 Coder by Alibaba — Benchmarks, Pricing & Review (202…

Qwen 3 Coder (Alibaba): Qwen 3 Coder by Alibaba.. 256K context. Pricing: from $0.00040/1K in. Specs, benchmarks and code examples.

What this reference supports

Qwen3-Coder is the Qwen team's open-weight agentic coding family, headlined by a large mixture-of-experts flagship (480B total, 35B active) released under Apache 2.0 with long-context support for repository-scale work. It targets teams building self-hosted coding agents, with MoE serving economics as the central implementation decision.

Qwen 3 Coder is attributed to Alibaba in ModelRefs' canonical registry. Tracked modalities: Text input and output. Primary use cases considered on ModelRefs: Self-hosted agentic coding assistants over private repositories; Repository-scale code understanding, generation, and repair pipelines.

This ModelRefs profile is Provisional and pending review — decision-support material, not a final or universal ranking. Confirm current behavior, access, pricing, limits, licensing, and lifecycle in Alibaba's own documentation, and evaluate Qwen 3 Coder on representative workloads before implementation.

Benchmark & Evaluation

ModelRefs does not yet hold qualifying sourced benchmark evidence for Qwen 3 Coder, so its benchmark coverage is incomplete. Treat any benchmark discussion as provisional and confirm results on representative workloads before selecting it.

  • No benchmark score is imported into this editorial record. Provider-reported evaluations support scoped notes only; canonical score records are governed separately with their own provenance.
  • The release post reports provider-run agentic coding evaluations; results are variant- and harness-specific.

Implementation considerations

  • Size infrastructure to the flagship's total MoE parameter count; smaller family variants trade capability for much lighter serving.
  • Preserve the documented chat and tool-calling templates; agentic coding quality is highly template-sensitive.
  • Open weights distributed via Hugging Face under Apache 2.0; hosted endpoints also exist on Alibaba Cloud and third-party platforms.
  • Map the exact artifact or endpoint version before comparing hosts; family variants differ substantially.

Risks and limitations

  • Open-weight results depend on the exact runtime, precision, quantization, and prompt template; reference results do not transfer automatically.
  • The release-specific license and acceptable-use policy must be reviewed before commercial deployment.

Source coverage

This reference is Provisional. Model behavior, access, pricing, limits, and lifecycle can change; verify the linked provider documentation and run task-specific evaluations before implementation.

Known coverage gaps:

  • Independent agentic-coding evaluation reproduction is not attached.
  • Variant-by-variant serving-cost and quality evidence is incomplete.

Sources

Continue your research

Use these connected ModelRefs sections to compare alternatives, inspect implementation paths, and review the evidence and governance boundaries relevant to Qwen 3 Coder by Alibaba — Benchmarks, Pricing & Review (202….