# diffusion-llm-mcp

> diffusion-llm-mcp — sandraschi-diffusion-llm-mcp. Use this tool when you need to accelerate batch inference for diffusion language models (dLLMs) with high-speed processing of approximately 200-400 tokens per second. It solves problems related to slow inference times and limited scalability in language model applications. The diffusion-llm-mcp tool takes in language model inputs and outputs rapid, HLE-shaped reasoning results, ideal for use cases requiring fast and efficient language processing.

Canonical page: https://skillsregistry.net/skills/sandraschi-diffusion-llm-mcp  
JSON: https://api.skillsregistry.net/v1/skills/sandraschi-diffusion-llm-mcp

## Description

FastMCP fleet MCP server for diffusion LMs (dLLM). DiffusionGemma on Goliath RTX 4090 — batch inference, HLE-shaped reasoning, ~200–400 tok/s. Doc phase; llama-diffusion-cli sidecar next. Complements local-llm-mcp.

## Trust

- **Trust score (0–1):** 0.89
- **Verification tier:** verified
- **Last scanned:** 2026-09-28

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** container
- **Runtime environment:** vm
- **License:** MIT
- **Updated:** 2026-09-28

## Source

- **Source listing:** [GitHub](https://github.com/sandraschi/diffusion-llm-mcp)

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "sandraschi-diffusion-llm-mcp"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/sandraschi-diffusion-llm-mcp` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/sandraschi-diffusion-llm-mcp/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt