@readmodelmag
MODEL

Best Tools · Coding

Tested

FlashProve

SummaryCPU verifier for inference proofs; our tests covered synthetic and toy statements only.

Filed November 5, 2026. Price, as published by the maker: Free, Apache-2.0. Research prototype; not maintained..

Open the maker’s site (opens in a new tab)

MODEL notes

URL: https://github.com/anthropics/FlashProve (opens in a new tab) · docs under docs/quickstart.md. We tested 0.16.0 (public tree 9e2bdd8…, tagged in-repo as the Oct 30 initial public release). What it does: Anthropic's research prototype for zero-knowledge proofs of AI inference. A provider commits to weights, proves a reply came from that commitment, and anyone checks the proof with cp_verify — on a normal CPU, without the weights. Pricing: Free, Apache-2.0. Research artifact; the README says it is not maintained and not accepting contributions. ✅ Tested hands-on: Linux x86_64, g++ 14 / make, no GPU, no model weights, no API key. Routes 1 and 2 from the quickstart only.

  • Honest proof: built bin/cp_verify in about 63 s (2.8 MB binary). Checking the shipped examples/accepted-proof (4.1 MB proof, synthetic small model) finished in about 1 s: exit 0, last line YES, result-v1 status=0 claim=exact-match for commitment ID 79a5a2bc…eb22183, tokens 61..63.
  • One flipped bit: XOR'd the middle byte of a copy of that proof. Exit 1, last lines: “an opening does not lead to its root” / NO: proof-refused.
  • Toy self-test: blocks/first_look.sh in about 40 s made six honest proofs (each ~576–578 KB) and tried 54 altered copies/calls: 0 accepted, 0 honest files refused. Honest copy ACCEPTED; bent copy rejected the same way.
  • The catch: these CPU routes prove the verifier and the proof library on synthetic/toy statements. They do not prove anything about a model anyone serves. The real Llama 4 Maverick path needs an H100, CUDA 13.2 Update 2, and on the order of hundreds of GB of host memory — we did not have that. For a real check, the commitment ID must come from a party you trust, not from files that arrived with the proof (Anthropic's own docs). We did not test GPU commit/prove, Maverick weights, the Lean soundness library, or run_suite.py. Verdict: If you care whether a “model receipt” can fail closed on a laptop CPU, this is the rare tool that actually rejects a bent proof in front of you. Treat it as a research verifier, not a production attestations product. Skip it if: you need a maintained, supported attestation service, or you only wanted the H100 Maverick demo and don't have the hardware.

Full issue: MODEL Nº 05.

  • Deslop

    Tested · Oxlint rules that fail common AI TypeScript slop patterns.

  • Ghostty 1.4.0

    Tested · The open-source terminal's biggest update in seven months. It also fixes a bug where a folder name from a cloned repo could quietly replace your clipboard.

  • Graphicle

    Tested · A local Rust binary that maps a repository into a code graph for a coding agent.