InferenceX (SemiAnalysisAI/InferenceX) is an open-source AI project on GitHub. Repository summary: Open Source Continuous Inference Benchmarking Qwen3.5, DeepSeek, GPTOSS - GB200 NVL72 vs MI355X vs B200 vs GB300 NVL72 vs H100 & soon™ TPUv6e/v7/Trainium2/3 Its focus includes evaluation and observability. It is suitable for extension, integration, and iterative delivery in real workflows.
License
Apache-2.0
Stars
893
Homepage
https://inferencex.com/Features
- Core capability: Open Source Continuous Inference Benchmarking Qwen3.5, DeepSeek, GPTOSS - GB200 NVL72 vs MI355X vs B200 vs GB300 NVL72 vs H100 & soon™ TPUv6e/v7/Trainium2/3
- Includes evaluation, tracing, or observability capabilities
- Repository: SemiAnalysisAI/InferenceX
- Primary language: Python
- Open-source license: Apache-2.0
- GitHub traction: about 893 stars
Use Cases
- Used for AI quality monitoring and regression evaluation
- Build internal AI workflow prototypes with InferenceX
- Validate InferenceX in production-like engineering scenarios
- Model evaluation and regression testing
- Monitoring AI application quality
- Business research and insight analysis
FAQ
Teams should first define integration boundaries and call patterns, then map repository capabilities into concrete interfaces, parameters, and access rules. GitHub repository: https://github.com/SemiAnalysisAI/InferenceX. Community traction is around 893 stars. License: Apache-2.0.
It usually works as an execution component or capability layer, with common deployment fits such as: Used for AI quality monitoring and regression evaluation, Build internal AI workflow prototypes with InferenceX, Validate InferenceX in production-like engineering scenarios.
