# QwenLM/Qwen3-Coder

Qwen3-Coder is the code version of Qwen3, the large language model series developed by Qwen team.

Repository: https://github.com/QwenLM/Qwen3-Coder
Canonical: https://ross.abutalabs.com/products/qwen3-coder
Language: Python
License Family: other
Last push: 2026-03-24T08:22:10+00:00

## Health v2 (maintenance only)
Score: 58/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 73, release rhythm 35, longevity 62
- inputs: {"age_days": 869, "days_push": 162, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 16810, forks 1245 (observed 2026-08-28T04:11:17.311090+00:00)

## What it is
Qwen3-Coder is a family of open-weight large language models specialized for coding and agentic coding tasks, including the 480B-A35B, 30B-A3B, and hybrid-attention MoE Qwen3-Coder-Next 80B variants. The repository provides model weights, quick-start inference examples, fill-in-the-middle support, and demos for building coding agents and local development workflows.

## Use cases
- run a local coding assistant LLM
- build an agentic coding tool that edits repos
- autocomplete code with fill-in-the-middle
- generate full websites from prompts
- fine-tune or evaluate a code LLM
- self-host an open-weight code model

## When to choose
- you need an open-weight code LLM you can self-host
- you want strong agentic coding performance without closed APIs
- you need a small MoE model for local development

## When to avoid
- you need a hosted API with SLA rather than model weights
- you lack GPU resources for large model inference
- you need a general-purpose chat model rather than a code specialist

## Facets
- artifact type: learning-resource
- maturity: active
- function: llm-inference, llm-training, agent-framework, code-review
- domain: large-language-models, artificial-intelligence, developer-tools
- platform: python, self-hosted, cross-platform
- tags: code-generation, open-weights, agentic-coding, moe, fill-in-the-middle, huggingface, coding-agent, ai-agents, gpu

## Member repositories
- QwenLM/Qwen3-Coder (main) score 58

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:17.311090+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:04:14.958375+00:00, confidence not recorded.
  - readme: https://github.com/QwenLM/Qwen3-Coder (fetched 2026-08-28T04:11:17.311090+00:00, sha 934d2e1e962c)
- Data as of 2026-08-30T08:39:29.467469+00:00.