964 9 months ago

Built from Unsloth's UD Q6_K_XL quant

tools
ollama run sammcj/devstral-small-24b-2505-ud:64k-q6_k_xl

Applications

Claude Code
Claude Code ollama launch claude --model sammcj/devstral-small-24b-2505-ud:64k-q6_k_xl
Codex
Codex ollama launch codex --model sammcj/devstral-small-24b-2505-ud:64k-q6_k_xl
OpenCode
OpenCode ollama launch opencode --model sammcj/devstral-small-24b-2505-ud:64k-q6_k_xl
OpenClaw
OpenClaw ollama launch openclaw --model sammcj/devstral-small-24b-2505-ud:64k-q6_k_xl

Models

View all →

Readme

GGUF: https://huggingface.co/unsloth/Devstral-Small-2505-GGUF/blob/main/Devstral-Small-2505-UD-Q6_K_XL.gguf

  • 128k tags default to 128k context which requires around 35GB vRAM.
  • 64k tags default to 64k context which requires around 27GB vRAM, num_batch set to 1024 (up from 512) for performance, you can tune num_batch to trade off speed and vRAM requirements.
  • cline tags are optimised for Cline & Roo Code Agentic Coding.