Local LLMs for coding
9 current open-weight models from 8 labs. Latest version of each family only.
Models
ModelLabParamsYour deviceScore
colgrep agent 2BLightOn2.52B2.3 GB ◇Mellum2.1 ThinkingJetBrains12.1B7.5 GB ◇AesCodeMicrosoft8.77B18.0 GB ◇Qwen 27B BlendJetBrains27.8B14.3 GB ◇Nex N2.5Nex-AGI35.1B66.6 GB ◇Nemotron Competitive CodingPreviewNVIDIA335B329 GB ◇Leanstral 1.5Mistral AI—113 GB ◇Kimi K2.7 CodeMoonshot AI1.03T568 GB ◇North Mini Code 1.0PreviewCohere30.5B58.1 GB ◇Newest first.
Min memory is the smallest listed option. ● stated by the publisher · ◇ estimate from file size, an 8K-token context and 0.5 GB for the runtime. — means no figure is available.