State of open code models (Sep 2025)
The best open code models on Hugging Face as of Sep 2025
Text Generation • 1T • Updated • 142k • • 2.32kNote Agent (https://docs.continue.dev/features/agent/)
Qwen/Qwen3-Coder-480B-A35B-Instruct
Text Generation • 480B • Updated • 75.2k • • 1.3kNote Agent (https://docs.continue.dev/features/agent)
zai-org/GLM-4.5
Text Generation • 358B • Updated • 43.5k • • 1.4kNote Agent (https://docs.continue.dev/features/agent/)
openai/gpt-oss-120b
Text Generation • 120B • Updated • 4.28M • • 4.55kNote Agent (https://docs.continue.dev/features/agent/)
zai-org/GLM-4.5-Air
Text Generation • Updated • 231k • • 584Note Agent (https://docs.continue.dev/features/agent/)
open-r1/OlympicCoder-32B
Text Generation • 33B • Updated • 23 • • 155Note Overall competitive coding (https://huggingface.co/blog/open-r1/update-3)
Qwen/Qwen3-Coder-30B-A3B-Instruct
Text Generation • Updated • 992k • • 962Note Agent (https://docs.continue.dev/features/agent/)
google/gemma-3-27b-it
Image-Text-to-Text • 27B • Updated • 1.49M • • 1.91kNote Chat (https://docs.continue.dev/features/chat)
mistralai/Devstral-Small-2507
Updated • 38k • 364Note Agent (https://docs.continue.dev/features/agent/)
openai/gpt-oss-20b
Text Generation • 22B • Updated • 7.22M • • 4.42kNote Agent (https://docs.continue.dev/features/agent/)
continuedev/instinct
Text Generation • 8B • Updated • 635 • 68Note Next Edit (https://docs.continue.dev/features/autocomplete/next-edit)
Qwen/Qwen2.5-Coder-7B
Text Generation • 8B • Updated • 234k • • 137Note Autocomplete (https://docs.continue.dev/features/autocomplete)
open-r1/OlympicCoder-7B
Text Generation • 8B • Updated • 292 • • 183Note Local competitive coding (https://huggingface.co/blog/open-r1/update-3)
Qwen/Qwen2.5-Coder-1.5B
Text Generation • Updated • 814k • • 83Note Autocomplete (https://docs.continue.dev/features/autocomplete)