inference-labs-inc/JSTprove

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

1306 stars · 15 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

JSTprove is a Rust CLI toolkit that generates zero-knowledge proofs of machine learning inference on ONNX models, built on Polyhedra Network's Expander GKR prover and the Expander Compiler Collection. It lets users prove that a model produced a given output without revealing model weights or inputs.

Use cases

• generate zero-knowledge proofs of onnx model inference
• verify ai model outputs without revealing weights
• prove neural network inference correctness on-chain
• zkml toolkit for verifiable machine learning
• prove conv and gemm model execution with zk proofs

When to choose

• you need verifiable, privacy-preserving proofs of ONNX model inference
• you want fast GKR-based zkML proving for supported ops like Conv, Gemm, and ReLU
• you need a CLI workflow to prove and verify ML inference

When to avoid

• your model uses ONNX ops outside the currently supported set
• you need a general-purpose ZK proof system rather than ML inference proofs
• you need a Python or hosted SDK instead of a Rust CLI

Facets

cli-tool · maturity active

machine-learning cryptography cli llm-inference machine-learning deep-learning developer-tools rust cli zkml zero-knowledge-proofs onnx verifiable-inference gkr-prover cryptography linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="inference-labs-inc/JSTprove")