DEDUPED Hot AI repos, turned into use cases Get the Monday 5

How to run Claude Code on a local quantized model with Unsloth

Use Unsloth start to launch Claude Code against a local GGUF model.

Remote script. The quick install pipes a script into sh; package managers are an alternative.

Who it is forA developer who wants a coding agent without cloud API usage.

The problemCoding agents default to cloud models.

The outcomeClaude Code running against a local model served by Unsloth.

What you need first

  • Windows, macOS or Linux, or Docker
  • A GPU for training

Steps

  1. Install Unsloth with curl -fsSL https://unsloth.ai/install.sh | sh
  2. Run unsloth start claude --model unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_XL
  3. Work in your repo
  4. Compare quality with your usual model

Prompt for your agent

Run `unsloth start claude --model unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_XL` and complete <small task>. Report response speed and whether the task succeeded.

Caveat. A 27B model at 4-bit needs substantial RAM or VRAM.

Use this in your AI agent

Ask your agent

Using the Deduped MCP, get the setup steps for unslothai/unsloth and "Run Claude Code on a local quantized model"

Add Deduped to Claude Code once

claude mcp add --transport http deduped https://dedupedai.com/mcp

Cursor, ChatGPT, others

Share: X LinkedIn Hacker News Email Copy link

Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer