Models / Run locally
Run Qwen3 Coder locally
Alibaba · 480B MoE (35B active) · open weights
Server-class: ≈280GB at 4-bit. It runs beautifully on a multi-GPU rig or a 512GB Mac Studio via MLX — for everyone else, Qwen3 Coder Flash is the local pick from the same family.
By quantization
Hardware requirements
| Quantization | ≈ File size | ≈ Memory needed | Runs on |
|---|---|---|---|
| Q4_K_M (4-bit) | 278 GB | 320 GB | Multi-GPU servers (4–8× 80GB) or 512GB Mac Studio clusters |
| Q8_0 (8-bit) | 514 GB | 591 GB | Multi-node GPU clusters — datacenter serving only |
| FP16 (full) | 1008 GB | 1159 GB | Multi-node GPU clusters — datacenter serving only |
Sizes are computed from the ≈480B parameter count with standard GGUF math and ~15% runtime overhead — verify against the model card before buying hardware. As a mixture-of-experts model only ≈35B parameters are active per token, which helps speed — but the full weights still have to fit in memory.
Local setup
How to run it
The shortest path is Ollama — it downloads a sensible default quantization and serves an OpenAI-compatible endpoint on localhost:
ollama run qwen3-coderFor serious throughput on server hardware, vLLM with tensor parallelism is the standard serving stack; llama.cpp and MLX (Apple silicon) win below that line.
The API route
Or skip the GPU entirely
The same model is one API call away at $0.9 / M input · $2.7 / M output — no download, no VRAM math, metered to the exact call. Free starter credits cover your first runs, and because the weights are open, nothing locks you in: start metered, move to your own hardware if sustained volume ever justifies it.
FAQ
Frequently asked questions
Can my GPU run Qwen3 Coder?
Qwen3 Coder is 480B MoE (35B active). At 4-bit quantization you need roughly 320GB of memory (multi-gpu servers (4–8× 80gb) or 512gb mac studio clusters). Server-class: ≈280GB at 4-bit. It runs beautifully on a multi-GPU rig or a 512GB Mac Studio via MLX — for everyone else, Qwen3 Coder Flash is the local pick from the same family.
Is there a Qwen3 Coder GGUF?
Open-weight releases in this family get community GGUF conversions on Hugging Face shortly after release — search the model name plus "GGUF" and pick the quantization your memory allows from the table above.
Is running Qwen3 Coder locally cheaper than the API?
Only at sustained volume. The API price is $0.9 / M input · $2.7 / M output with no hardware, electricity, or ops cost — a machine that can serve this model costs more per month idle than most teams' entire inference bill. Prototype metered, self-host when utilization justifies it.
Can I use the outputs commercially?
Apache-2.0 open weights; commercial use permitted.
Explore
Run other models locally
One key. Every model. Exact prices.
Prototype on Qwen3 Coder with free starter credits while the weights download. Free starter credits included — no subscription required.