Zcash (ZEC) fell by about 5.9% in 24 hours to roughly $1,190 as the wider crypto market also experienced a drop. Amidst the decline, Helius founder Mert Mumtaz pointed out that Anthropic’s ne
Zcash (ZEC) fell by about 5.9% in 24 hours to roughly $1,190 as the wider crypto market also experienced a drop.
Amidst the decline, Helius founder Mert Mumtaz pointed out that Anthropic’s new FlashProve tool uses the same type of privacy math that Zcash helped bring to the world.
Zcash (ZEC) fell by about 5.9% in 24 hours and is currently trading at $1,190. The whole market declined, with Bitcoin (BTC) trading near $76,673, and the total crypto market value falling by about 1.8%.
In those same 24 hours, Helius founder Mert Mumtaz said in a post on X that “anthropic/claude are now using the same cryptography that zcash was first in the world to pioneer,” and ended with “probably nothing.” The post drew a bit of attention, gaining more than 1,000 likes and over 170 replies at the time of writing.
In May, Mumtaz went on the Bankless podcast and argued that Zcash is worth more than the market gives it credit for, especially when compared with XRP and BNB. He said ZEC’s value should be a share of Bitcoin’s total market value; about 10% has been suggested.
Zcash grew out of the 2014 paper “Zerocash: Decentralized Payments from Bitcoin.” In 2024, the IEEE Symposium on Security and Privacy gave it a Test of Time Award. The committee said the paper gave “an actual example for how to use zk-SNARKs in practice.” The Zcash network launched in October 2016.
One person, Alessandro Chiesa, is tied to both projects. He co-wrote the Zerocash paper and now works at EPFL. Chiesa is also linked to StarkWare and is on the FlashProve contributor list with Anthropic’s Roy Rinberg and Alex Allain.
This is not the first Anthropic and Zcash overlap this year. Zcash founder Zooko Wilcox said in June that an audit by Anthropic’s Mythos AI, requested by Shielded Labs, found no serious new bugs in the protocol.
Notably, the Zcash community is currently arguing over its roughly $125 million grants treasury.
What is FlashProve, and how costly is it to run?
Anthropic released FlashProve, an open-source research prototype, on November 2. It uses a hash-based zkSNARK to prove that an answer came from a specific AI model, without showing the model’s inner settings, known as weights.
FlashProve aims to close a trust gap. When a user sends a prompt to an AI service, they cannot be sure the reply came from the model they paid for. Before the answer is delivered, a routing service could swap in a cheaper model to respond, or the reply could be changed along the way. The FlashProve tool is designed for mixture-of-experts models, which send each input through a few specialized parts of the system.
FlashProve works in two steps. First, the provider publishes a small commitment of about 6.5 MB that locks in one exact set of weights while keeping them secret. This takes about 39 minutes. Second, for each prompt and reply, the system creates a short proof that the committed model produced those exact output tokens.
Anthropic tested FlashProve on Llama 4 Maverick, an AI model from Meta (NASDAQ: META). The model has 400.7 billion parameters and was run in a compressed 4-bit form. The test used one NVIDIA (NASDAQ: NVDA) H100 chip, which handled about 110 tokens per second. A token is a small piece of text, such as a word or part of a word.
The system handled 24 requests in about 223 seconds. Each request was 1,024 tokens long and produced a proof of about 19 MB. Checking a proof took only 17 to 21 seconds on a regular computer chip (CPU), and the checker did not need access to the model’s weights.
However, compared with vLLM inference, proof generation took about 72 to 94 times more time per GPU, depending on load.
Anthropic calls FlashProve an intermediate research tool, not production software. The GitHub page labels the code an unmaintained research artifact under the Apache 2.0 license. The project also says it has not had an independent security audit.
The smartest crypto minds already read our newsletter. Want in? Join them.
