Best open coding models to run locally
Coding models are language models trained or fine-tuned on code. They can generate, complete, explain, and debug code across multiple programming languages. Run them locally with the same tools as other language models.
Last updated 3 November 2026 · refreshed after every Hugging Face sync · 9 models
| # | Model | Size | Memory | License | Downloads (30 days) | Likes |
|---|---|---|---|---|---|---|
| 1 | Qwen3-Coder-30B-A3B-Instruct Qwen | 30B · FP16 | ~72 GB | Apache 2.0 | 1.1M | 1.3K |
| 2 | Qwen2.5-Coder-7B-Instruct Qwen | 7B · FP16 | ~17 GB | Apache 2.0 | 1.4M | 859 |
| 3 | Qwen3-Coder-30B-A3B-Instruct-GGUF unsloth | 30B · GGUF | ~18 GB | Apache 2.0 | 1.1M | 1.2K |
| 4 | Qwen2.5-Coder-14B-Instruct Qwen | 14B · FP16 | ~34 GB | Apache 2.0 | 2M | 211 |
| 5 | Qwen3-Coder-30B-A3B-Instruct-FP8 Qwen | 30B · FP16 | ~72 GB | Apache 2.0 | 1.1M | 240 |
| 6 | Qwen3-Coder-Next-FP8 Qwen | – · FP16 | – | Apache 2.0 | 812.8K | 212 |
| 7 | Qwen2.5-Coder-32B-Instruct-AWQ Qwen | 32B · AWQ | ~20 GB | Apache 2.0 | 1M | 50 |
| 8 | Qwen3-Coder-30B-A3B-Instruct-AWQ-4bit cyankiwi | 30B · AWQ | ~18 GB | Apache 2.0 | 777.2K | 67 |
| 9 | gemma-4-12B-coder-fable5-composer2.5-v1-GGUF yuxinlu1 | 12B · GGUF | ~8 GB | Apache 2.0 | 110.7K | 3K |
How we rank
Text-generation models whose name or metadata signals coding capability. We include models seen on Hugging Face in the last 14 days with at least 10,000 downloads in the last 30 days. Score = log10(30-day downloads) + ½ × log10(likes + 1), so popularity leads and likes break near-ties. We keep one entry per model name and leave out adult and uncensored builds. Memory is estimated from parameter count and the published file format (16-bit ≈ 2 bytes per parameter, 4-bit ≈ 0.5) plus 20% headroom; long contexts need more. A quantised copy of a bigger model may also fit — check the model page.