AlphaUncommon
StartLuxLabs/StartLux-Prism
Fast, exact inference for StartLux-Decision models on NVIDIA GPUs: the released GGUF files on int8 and 4-bit tensor cores, behind the /v1/systemone API, with up to 5.3× lower latency than vLLM in FP8 and bit-identical answers.
- Owner
- StartLuxLabs (organization)
- Stars
- Stars: 270
- Forks
- Forks: 5
- Language
- Python
- License
- None
- Created
- 2026
- Topics
- calibrationcudacuda-graphsdecision-makingggufh200inference-engineint8llm-inferencenvidia
- Tier
- Alpha (250 to 699 ★) What this tier means
More repos by StartLuxLabs
All 3 repos by StartLuxLabs- StartLuxLabs/StartLux-DecisionStars: 1,644PythonBeta
- StartLuxLabs/DecisionPivotStars: 152PythonPrototype
Nearby in Python
All Python repos- arsumbris/arsumbrisStars: 270PythonAlpha
- NEORUAA/HyperOS-AVDStars: 270PythonAlpha
- 10291124jcjc-dotcom/ai-anime-seriesStars: 270PythonAlpha
- hiidef/oauth2appStars: 270PythonAlpha
- kvh/recurrentStars: 270PythonAlpha
- deepgully/meStars: 270PythonAlpha