Best Coding Models You Can Run Locally

Open-source coding models have caught up to commercial options for most day-to-day tasks. These are the strongest local models for code completion, generation, and review — sorted by Hugging Face download counts as a proxy for community adoption.

  1. 1

    all-MiniLM-L6-v2

    0.023B params

    Tiny embedding model. Only 23MB. Perfect for on-device search.

    Min VRAM: 0.1GBQuant: Q8_0Size: 0.023GBLicense: apache-2.0
  2. 2

    Nomic Embed Text v1.5

    0.137B params

    High quality text embedding model. 137M params. Good for RAG and search.

    Min VRAM: 0.3GBQuant: Q8_0Size: 0.139GBLicense: apache-2.0
  3. 3

    TinyLlama 1.1B

    1.1B params

    Lightweight 1.1B chat model based on Llama architecture. Great for phones.

    Min VRAM: 1.12GBQuant: Q4_K_MSize: 0.623GBLicense: apache-2.0
  4. 4

    Moondream 2

    1.8B params

    Ultra-compact vision model. Only 1GB. Answers questions about images.

    Min VRAM: 1.5GBQuant: Q4_K_MSize: 1GBLicense: apache-2.0
  5. 5

    Qwen 2.5 Coder 7B

    7.6B params

    Strong 7B code model rivaling larger coding models. Excellent for local development.

    Min VRAM: 4.86GBQuant: Q4_K_MSize: 4.361GBLicense: apache-2.0
  6. 6

    Phi-3.5 Vision

    4.2B params

    Vision-language model from Microsoft. Can understand images and documents.

    Min VRAM: 3.2GBQuant: Q4_K_MSize: 2.5GBLicense: mit
  7. 7

    Phi-4 Mini 3.8B

    3.8B params

    Latest Phi mini with strong reasoning. Drop-in upgrade from Phi-3.5 Mini.

    Min VRAM: 2.82GBQuant: Q4_K_MSize: 2.321GBLicense: mit
  8. 8

    Qwen 2.5 Coder 14B

    14B params

    Powerful 14B code model. Excellent for complex programming tasks.

    Min VRAM: 8.87GBQuant: Q4_K_MSize: 8.371GBLicense: apache-2.0
  9. 9

    Phi-3.5 Mini 3.8B

    3.8B params

    Tiny but capable 3.8B model. Runs on almost any hardware including phones.

    Min VRAM: 2.73GBQuant: Q4_K_MSize: 2.229GBLicense: mit
  10. 10

    Phi-4

    14B params

    Microsoft's 14B parameter model. Punches well above its weight on reasoning.

    Min VRAM: 8.93GBQuant: Q4_K_MSize: 8.431GBLicense: mit
  11. 11

    Qwen 2.5 Coder 1.5B

    1.5B params

    Compact code model with solid code generation and understanding abilities.

    Min VRAM: 1.54GBQuant: Q4_K_MSize: 1.041GBLicense: apache-2.0
  12. 12

    Qwen 2.5 Coder 0.5B

    0.5B params

    Smallest code model. Default code assistant - runs on any iPhone. Great for code completion and simple programming tasks.

    Min VRAM: 1.13GBQuant: Q8_0Size: 0.629GBLicense: apache-2.0
  13. 13

    EXAONE 3.5 7.8B

    7.8B params

    7.8B model from LG. Strong bilingual Korean/English.

    Min VRAM: 4.94GBQuant: Q4_K_MSize: 4.443GBLicense: other
  14. 14

    Qwen 2.5 Coder 3B

    3B params

    Capable 3B code model. Good balance of coding ability and resource usage.

    Min VRAM: 2.46GBQuant: Q4_K_MSize: 1.96GBLicense: apache-2.0
  15. 15

    MiniCPM-V 2.6

    2B params

    Efficient multimodal model with strong image understanding. Optimized for edge devices.

    Min VRAM: 2.1GBQuant: Q4_K_MSize: 1.6GBLicense: apache-2.0

Related

All coding models