Optimized GPU compiler for LLM inference. Choose from a list of optimized recipes or optimize your own model via kernel fusion, autotuning, and advanced scheduling. Run benchmarks across different GPU types and configurations, track results and share experiments with the community.
Latest release v0.3.22 · 11 Oct 2026
10 files for Claude Code, Codex and OpenCode: discover-models, onboard-model, release-serving-image, reproduce-article-benchmarks and 6 more — 8,916 tokens loaded in every session.
AGENTS.md A 8,911 tok CLAUDE.md A 5 tok .agents/skills/discover-models/SKILL.md A 77 tok .agents/skills/onboard-model/SKILL.md A 84 tok .agents/skills/release-serving-image/SKILL.md C 94 tok .agents/skills/reproduce-article-benchmarks/SKILL.md A 95 tok .agents/skills/run-experiment/SKILL.md A 87 tok .agents/skills/start-remote-server/SKILL.md C 61 tok .agents/skills/refresh-golden/SKILL.md A 115 tok .claude/agents/architect.md A 44 tok These files are cloudrift-ai/emmy's own configuration — they tell Claude Code, Codex and OpenCode how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.