by raullenchai
The fastest local AI engine for Apple Silicon. 4.2x faster than Ollama, 0.08s cached TTFT, 100% tool calling. 17 tool parsers, prompt cache, reasoning separation, cloud routing. Drop-in OpenAI replacement. Works with Claude Code, Cursor, Aider.
# Add to your Claude Code skills
git clone https://github.com/raullenchai/Rapid-MLXLast scanned: 5/6/2026
{
"issues": [],
"status": "PASSED",
"scannedAt": "2026-05-06T06:29:36.289Z",
"semgrepRan": false,
"npmAuditRan": true,
"pipAuditRan": true
}Rapid-MLX is an open-source testing skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by raullenchai. The fastest local AI engine for Apple Silicon. 4.2x faster than Ollama, 0.08s cached TTFT, 100% tool calling. 17 tool parsers, prompt cache, reasoning separation, cloud routing. Drop-in OpenAI replacement. Works with Claude Code, Cursor, Aider. It has 3,251 GitHub stars.
Yes. Rapid-MLX passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.
Clone the repository with "git clone https://github.com/raullenchai/Rapid-MLX" and add it to your Claude Code skills directory (see the Installation section above).
Rapid-MLX is primarily written in Python. It is open-source under raullenchai on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other Testing skills you can browse and compare side by side. Open the Testing category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh Rapid-MLX against similar tools.
No comments yet. Be the first to share your thoughts!
Top skills in this category by stars
1. Install (one command, detects your RAM, picks a starter model):
curl -fsSL https://rapidmlx.com/install.sh | bash
Installs Python 3.10+ if missing, creates an isolated venv at ~/.rapid-mlx/, symlinks the rapid-mlx CLI into ~/.local/bin/, and prints a serve command sized to your Mac (8–23 GB → qwen3.5-4b-4bit; 24–47 GB → gpt-oss-20b-mxfp4-q8; 48–95 GB → qwen3.6-35b-8bit; 96 GB+ → gpt-oss-120b-mxfp4-q8).
curl | bashsecurity.install.shis served over HTTPS (HSTS-preload) fromrapidmlx.comand is a byte-identical mirror ofinstall.shat the current release commit — read it before running if you like. Two verified alternatives:
- Pin to a commit hash —
curl -fsSL https://raw.githubusercontent.com/raullenchai/Rapid-MLX/<commit>/install.sh -o install.sh && shasum -a 256 install.sh && bash install.sh- Skip the shell script entirely — use Homebrew,
uv, orpipbelow.
See Alternative install methods for the non-curl paths.
2. Chat with a model right now:
rapid-mlx chat
Defaults to qwen3.5-4b-4bit. First run downloads the weights (~2.5 GB) with a progress bar and drops you into a REPL. Type /help for slash commands, /exit to quit.
3. Or serve it for use from other apps:
rapid-mlx serve qwen3.5-4b-4bit
Starts an OpenAI-compatible HTTP server bound to http://localhost:8000. Point any OpenAI SDK / client (Cursor, Aider, LangChain, OpenCode, PydanticAI, your own scripts) at http://localhost:8000/v1; Claude Code / Anthropic SDK uses http://localhost:8000 (the Anthropic messages route lives at /v1/messages under the same host).
curl http://localhost:8000/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{"model":"default","messages":[{"role":"user","content":"Say hello"}]}'
from openai import OpenAI
client = OpenAI(base_url="http://localhost:8000/v1", api_key="not-needed")
print(client.chat.completions.create(
model="default",
messages=[{"role": "user", "content": "Say hello"}],
).choices[0].message.content)
Vision / audio / diffusion models? Base install is text-only (~460 MB). Vision, audio, embeddings, and DFlash speculative decoding ship as opt-in extras. → Optional extras
Not into the terminal? Rapid-MLX Desktop bundles the same engine inside a one-click Mac app.
| Apple-Silicon-native | Pure MLX kernels — no llama.cpp fallback, no Metal shim. Continuous batching, prompt cache (radix + DeltaNet RNN snapshots), and TurboQuant K8V4 KV codec run at native MLX bandwidth on M1 → M4. |
| Drop-in OpenAI / Anthropic API | /v1/chat/completions, /v1/responses (Codex CLI), /v1/messages (Anthropic SDK / Claude Code), /v1/embeddings, /v1/audio/* — same wire as ChatGPT / Claude, no client adapter. |
| Tier-1 ecosystem coverage | 8 agent CLIs and 3 Python frameworks are wire-verified against real weights every release — Codex CLI, Claude Code, OpenCode, Qwen Code, OpenHands, Hermes Agent, Aider, Kilo Code + LangChain, PydanticAI, smolagents. |
| Chat in the terminal | rapid-mlx chat qwen3.5-9b-4bit |
Streaming REPL, /help for slash commands, --think / --no-think to control CoT. |
| OpenAI server for your apps | rapid-mlx serve qwen3.5-9b-4bit |
Point Cursor, Aider, LibreChat, Open WebUI, LangChain at http://localhost:8000/v1. |
| Agent backends | rapid-mlx serve qwen3.6-35b-8bit &rapid-mlx agents codex --setup && codex |
8 Tier-1 agents auto-configure once the server is up — see Tier-1 support. |
| Benchmark your Mac | rapid-mlx bench qwen3.5-9b-4bit --submit |
Standardized B=1 bench, opens a PR to publish your row on rapidmlx.com. |
→ One-shot IDE setup with rapid-mlx launch <cursor|claude-code|cline|continue-dev>
Every row below has a rapid-mlx agents <name> --setup config template (except Claude Code, which is one env-var) and an integration test that drives the same wire the real client drives against a live server.
| Agents (8) | Frameworks (3) |
|---|---|
| Codex CLI · Claude Code · OpenCode · Qwen Code · OpenHands · Hermes Agent · Aider · Kilo Code | LangChain (+ LangGraph) · PydanticAI · smolagents |
Also compatible with any OpenAI-compatible client via http://localhost:8000/v1 — Cursor, LibreChat, Open WebUI, and more plug in with a single URL change.
→ Full 8×3 agent matrix + 3×3 framework matrix (test cells + xfail reasons) → Codex CLI · Claude Code · OpenCode · Qwen Code · OpenHands · Hermes · Aider · Kilo Code
The installer's RAM detector picks a sensible default. If you want to shop the full catalog: rapid-mlx models lists every alias, rapid-mlx info <alias> shows the per-alias profile (parser, MoE / hybrid flags, KV codec eligibility, speculative-decoding gates).
| RAM | Recommended | One-shot |
|---|---|---|
| 8–23 GB MacBook Air/Pro | qwen3.5-4b-4bit |
rapid-mlx serve qwen3.5-4b-4bit |
| 24–47 GB MacBook Pro / Mac Mini | gpt-oss-20b-mxfp4-q8 |
rapid-mlx serve gpt-oss-20b-mxfp4-q8 |
| 48–95 GB Mac Studio | qwen3.6-35b-8bit |
rapid-mlx serve qwen3.6-35b-8bit |
| 96 GB+ Mac Studio / Pro | gpt-oss-120b-mxfp4-q8 |
rapid-mlx serve gpt-oss-120b-mxfp4-q8 |
→ Full RAM tier map + serve flags per tier → Every alias, quant, and family (128+ aliases across 30+ families) · interactive at models.rapidmlx.com
The curl one-liner above wraps all of these — reach for these only if you already manage Python yourself.
brew tap raullenchai/rapid-mlx
brew trust raullenchai/rapid-mlx
brew install rapid-mlx
Upgrade with brew upgrade rapid-mlx. If brew install stalls on Tapping homebrew/core, run brew tap homebrew/core --force once (one-time ~1.3 GB download) and retry.
uv tool install rapid-mlx@latest
Don't have uv yet? curl -LsSf https://astral.sh/uv/install.sh | sh. Upgrade with uv tool upgrade rapid-mlx.
python3.12 -m pip install rapid-mlx
If pip install rapid-mlx says "no matching distribution", your Python is too old. brew install python@3.12 first. Upgrade with pip install -U rapid-mlx.
For image-input / VLM models (Qwen-VL, true multimodal), install the vision extra: pip install 'rapid-mlx[vision]' — see Optional extras.