TensorCash (@Tensorcash) header
TensorCash (@Tensorcash) profile picture

The verification and settlement layer for global AI inference.

Mainnet
Joined June 2026
Photo by TensorCash (@Tensorcash)Photo by TensorCash (@Tensorcash)Photo by TensorCash (@Tensorcash)Photo by TensorCash (@Tensorcash)Photo by TensorCash (@Tensorcash)Photo by TensorCash (@Tensorcash)Photo by TensorCash (@Tensorcash)Photo by TensorCash (@Tensorcash)Photo by TensorCash (@Tensorcash)Photo by TensorCash (@Tensorcash)
TensorCash v1.2.5 is now available. It adds block history pricing. The work a block needs depends on earlier blocks from the same payout address, to reward higher-quality inference. It starts at block 36,875 (~Fri 20 Nov, 12:00 UTC). Block producers must upgrade before then.
1
1
18
631
Release notes and downloads: git.tensorcash.org/tensorcas… Running 1.2.5-rc1? Upgrade before block 36,750. Sovereign verifiers need ruleset 7 support before block 37,375. The new pricing applies to Qwen3-8B and Qwen3.8-27B and phases in over 1,000 blocks, with full pricing from 37,875.
1
156
About 15% of the circulating $TSC supply sits on centralised exchanges. After the record $6.3B hack on Binance, our advice is to keep your coins in a self-custodial wallet, where you hold the keys. TensorCash Core is the official wallet and runs on macOS, Windows and Linux.
3
2
23
1,307
TensorCash Core and the third-party wallets are listed here: tensorcash.org/wallets If you use a web wallet, the official browser extension checks its code against signed releases.
3
292
Useful computation produces proof, and proof produces value. Without proof, the quality of an answer can’t be checked, so its value is uncertain. A miner on TensorCash runs AI inference and records a proof of the model it used. With that proof, the inference can be sold as verified, and every eligible window is a chance at the block reward.
5
6
31
1,743
Mining on TensorCash produces AI inference, with a proof of the model behind it. Here is how that proof is checked, and why faking it never pays.
3
10
3
37
6,342
Mining should produce something people actually need. That is the first principle behind the TensorCash protocol. The work that secures the chain is AI inference, so the electricity spent on mining produces answers people can use, each with a proof of the model that wrote it.
4
10
46
1,598
TensorCash v1.2.4b is now available. It introduces ruleset 6, with dense error checking, historical content non-reuse checks and repetitive entropy filters, to reward higher quality blocks and reduce fork risk. It activates at mainnet block 35,300 (~Tue 10 Nov, 12:00 UTC). Block producers need to upgrade before then, and verifiers should update to support ruleset 6. Until that height, v1.2.4 keeps working.
3
4
29
1,398
A miner on the TensorCash network wrote a website for a coding agent on Qwen3.8-27B, and received 0.01136 TSC for this inference. Inference and a shot at the block reward, from the same unit of electricity. A live run by @boblabs_eu, an independent team building on the network.
12
8
5
46
4,476
Every miner on the network creates inference like this while competing for the block reward, whether it gets sold or not. Platforms such as @boblabs_eu, @AskTillyAI and @inferenceFI are bringing that supply together with demand.
1
10
580
Network difficulty on TensorCash reached a new all-time high today. More AI inference is running to secure the chain than ever before, and every block still comes with a proof of the model that produced it.
6
10
5
57
2,936
Great to see verifiable inference getting the attention it deserves. Tensorcash is the layer-1 where it is running in production, today.
Anthropic is releasing an open source library, FlashProve, for provable inference. It was measured on Llama 4 Maverick, a 400B-parameter mixture of experts model. Code: github.com/anthropics/FlashP… Blog post: anthropic.com/research/prova…
2
7
1
36
1,451
On TensorCash, the work that secures the chain is AI inference. Miners run AI models, and every answer they write is also an attempt at the next block. One computation, two outputs: a secured chain and inference that someone actually uses.
3
15
2
53
4,257
Replying to @drwconvexity
On the issue of sequencing @Tensorcash’s DEX (in development) addresses this in a very innovative way. Validators operate in a distributed network topology, and speed of light + order info cryptographic fragmentation guarantee stochastically fair sequencing.
6
1
19
646
On TensorCash, two people can trade directly, with no one else in between. The funds stay in their own wallets until the trade settles, removing the risk of an exchange losing or freezing them. This video follows one of those trades, from the offer to the block it settled in. One unit of a call option on the network’s difficulty was traded for 0.9 TSC. The option and the TSC moved in one transaction, straight from one wallet to the other. Each trade on the protocol is trustless, and neither side can end up paying without receiving. All in an ordinary transaction, with no escrow, no multisig and no trace of the negotiation, which keeps it cheap and private.
3
12
4
46
4,049