JSTprove Public Launch Brings Verifiable AI to Developers

October 31, 2025 – Inference Labs has announced the JSTprove public launch, a major step toward making Zero-Knowledge Machine Learning (zkML) accessible to developers and organizations worldwide. Built on Polyhedra Network’s Expander backend, JSTprove simplifies the generation and verification of AI inference proofs through an easy-to-use command-line interface, bridging the gap between cryptography and machine learning.

JSTprove Public Launch Redefines zkML Accessibility

The JSTprove public launch introduces a full end-to-end pipeline for zero-knowledge AI proofs. It accepts ONNX models and automates proof generation and verification through Polyhedra’s Expander Compiler Collection. Users can generate circuits, witnesses, and proofs without cryptographic expertise, making verifiable AI practical for real-world applications in finance, cybersecurity, and healthcare.

The toolkit supports core CNN layers such as Conv2D, MaxPool, ReLU, and Fully Connected, and it’s fully open-source allowing developers to inspect, audit, and extend its capabilities.

Inside JSTprove: How Verifiable AI Works

At its core, JSTprove turns conventional AI inference into transparent, auditable computation. The process involves quantizing an ONNX model, generating circuits via Expander, and creating proofs that confirm the model executed as intended without exposing private parameters.

This ensures trust in AI decisions, preserving both accuracy and confidentiality. Each command outputs reproducible artifacts for traceability, making JSTprove a true bridge between AI and cryptographic proof systems.

Future Roadmap for JSTprove Public Launch

Following the JSTprove public launch, Inference Labs plans to expand support for models such as ResNet and Transformers, introduce GPU acceleration via zkCUDA, and integrate recursive proof composition for multi-layer verification. The team also invites open-source contributions, encouraging developers to optimize circuits, add new model types, and participate in the project’s community channels on GitHub, Telegram, and Twitter.

With its modular design and transparent methodology, JSTprove paves the way for a new era of verifiable AI systems where performance and provability coexist seamlessly.

The Talk

Real voices. Real reactions.

KOLsMediaCommunity

LOW

@inference_labs Real value doesn’t shout, it builds,let’s see if this one can prove that. 🔥

"https://t.co/PGdv9IScSy"— @

Jamal

@Jam170310

1 day ago

LOW

@inference_labs DSperse makes zkML efficient with slice-based verification.

"https://t.co/PGdv9IScSy"— @

Gbemi

@Gbemi9949977769

1 day ago

LOW

@inference_labs zkML is one of the most promising directions in AI privacy and trust, but it has long remained confined to labs and research prototypes

"https://t.co/PGdv9IScSy"— @

XSTIANO/ FINC / $SGi

@XSTIANO17887

1 day ago

Do you want to read and know more about Altcoins? Check out our Altcoin Arena for sentiment analysis, latest updates, community reactions, price predictions and project reviews on the best altcoins in crypto.