AI article
Anthropic's FlashProve Proves Which Model Answered You, at 70-95x vLLM's GPU Time
Community description: Anthropic open-sourced a prototype that proves a reply came from committed weights. It runs on Llama 4 Maverick, not Claude, in a 1,024-token window.
Dev.to | Nov 3, 2026 | Gabriel Anhaia
Automated excerpt
The provider publishes a commitment ID bound to one set of weights. The demo model is Llama 4 Maverick, a 400B-parameter mixture-of-experts model with about 17B parameters active per token, using public 4-bit weights. Commitment: about 39 minutes, once per model, good for about 16 million requests before the provider recommits.
Selected automatically from source text; not independently written or fact-checked. Read the original for full context.