RapidMlx Studio
Local LLM server studio for Apple Silicon
RapidMlx Studio — run coding agents on a private local LLM server with fast Apple Silicon inference plus live benchmarks
How to use RapidMlx Studio
- Press Start server on the dashboard to bring the local endpoint online
- Open Models and download then load the model you want to serve
- Pick a coding agent on the Agents tab and launch it against the server
- Open the Playground, type a prompt and watch the streamed reply with tok per s
- Check Benchmarks to compare decode speeds across engines
Features
- Server control panel with start and stop and OpenAI plus Anthropic compatibility badges
- Model library with download progress, load and unload per model
- Chip profiles with estimated decode speed in tokens per second
- Agent launcher for 7 coding agents with copyable launch commands
- Streaming playground chat with live tokens per second meter
- Temperature, max tokens and context sliders with validation
- Benchmark table comparing Rapid-MLX, oMLX and LM Studio decode speeds
Frequently asked questions
Is it free?
Yes, unlimited.
Login to use this free online tool. Create a free account to get started.
