anthropics/FlashProve
A research prototype for zero-knowledge proofs of AI inference. Not maintained.
From the publisher
README & documentation
Source previewRead the project’s overview, installation instructions and usage examples. The original README is the source of truth.
Version 0.16.0 · A research artifact: not maintained, not accepting contributions. · Apache 2.0 FlashProve lets an AI provider prove which model answered a request, without revealing the model's weights. one set of weights. The weights themselves stay secret: every published value that depends on them is masked, except a declared list. model's output. weights. FlashProve runs Llama 4 Maverick at full size, with a 1,024-token context window, on a single NVIDIA H100. The prover reaches about 110 tokens/second of prefill at 24 requests per batch (Performance). The protocol's soundness is proved in Lean 4. A provider that tries to pass off output that is not the committed model's is caught, except with probability about q/2^124 + 2^-101 after q hash queries. lean/THEOREM.md has the theorem and how to check it. Building FlashProve, running it and checking its output take long builds and many exact comparisons. That work is meant…
Read the full README ↗ · Preview checked 2026-11-04T13:05:49.827Z
Inside the original README — Document outline
- FlashProve
- What it is
- What model it runs
- What is proved
- Built to be run by an agent
- For agents
- Hardware and software
- 1. Build the verifier and check a stored proof
- 2. Toy proofs (CPU)
- 3. Prove on a GPU
- Verifying a real proof
- Tests
Headings are captured from the source. Links open the publisher’s document, not a locally hosted copy.
Links from the README
References supplied by the publisher, not independently verified endorsements. Check the destination before downloading files or entering credentials.
Documentation belongs to its respective authors. Reported project/model license: Apache-2.0. A listing is not a grant of reuse or training rights. Confirm the document’s own terms at the source.
What this repository does
A research prototype for zero-knowledge proofs of AI inference. Not maintained.
Repository facts
- Owner
- anthropics
- Primary language
- Lean
- Stars
- 670
- Forks
- 57
- Open issues + pull requests
- 2
- License
- Apache-2.0
- Archived
- No
- Default branch
- main
- Created
- 2026-10-30T19:29:59.000Z
- Last push
- 2026-10-30T21:00:23.000Z
Topics and intended use
No topics were included in the latest source metadata.
Review the README for scope, installation, examples and limitations. We do not run repository code or certify it.
Evaluate before installing
Review licensing and dependencies, inspect recent commits and unresolved issues, and test in an isolated environment before production use. Stars and forks alone cannot answer these questions.
Source and freshness
Source: GitHub. Metadata observed 2026-11-04T12:18:51.791Z. Daily imports are snapshots, not real-time monitoring.
Popularity and source listings do not establish security, suitability, licensing rights or benchmark performance.