- Category
- Developer Tools
- Rank
- No. 2666Tools index
- Pricing
- Open Source
- Type
- TOOL
- GitHub
- 5 stars
- Latest release
- v1.1.4
- Date
About
Pulsar is an inference engine for running Qwen3.8-27B and Qwen3.6-35B-A3B locally on Apple Silicon, built on Metal GPU kernels and speculative decoding. It exposes an OpenAI-compatible API. The README reports up to 452 tok/s for code editing with Qwen3.8-27B, 1.36x faster than lithos-metal on a 10-prompt greedy test.
Tags
inferenceapple-siliconqwenmetallocal-llmspeculative-decoding
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.