Table of Contents
BlueTAO has launched an alpha version of a consumer AI chat interface built exclusively on Bittensor subnets, routing inference through Chutes (SN64) and targeting users who want private, uncensored AI access without a cryptocurrency wallet or technical setup.
This is in Alpha only. But a $TAO -powered ChatGPT-like interface. https://t.co/Aha5LWUjJt . Only using bittensor subnets (like @chutes_ai ) for text and image inference. (cc @const_reborn , @BarrySilbert ). Any suggestions or bugs welcome. Will be 1/250th the price of chatgpt.
— James Altucher (@jaltucher) May 22, 2026
`
The product supports persistent login, meaning conversation history carries across sessions. Users can upload documents and ask questions about them. Inference runs on DeepSeek V3.2, an open-weight large language model developed by DeepSeek, served through Chutes' distributed network of GPU miners on SN64. The team has disclosed that the product is in early testing and subject to errors and changes.
BlueTAO positions itself as a consumer-facing alternative to centralized AI products like ChatGPT, drawing a direct comparison to Venice.ai, the privacy-first AI interface that reached 1.3 million registered users and 50,000 daily active users by early 2026 by routing inference through its own decentralized GPU network. Where Venice built and operates its own infrastructure, BlueTAO routes through Bittensor's existing subnet economy.
The cost structure follows directly from how Bittensor allocates compute. Miners on SN64 earn TAO emissions for contributing GPU hardware to the network, and that subsidy flows through to end users as inference priced well below centralized alternatives. Chutes has reported up to 85% lower costs than traditional cloud inference providers such as AWS. DeepSeek V3.2 through DeepSeek's own API costs approximately $0.14 per million input tokens and $0.28 per million output tokens. ChatGPT Plus runs $20 per month. BlueTAO has not yet confirmed its own pricing model publicly.
James Altucher plugged a front-end onto Bittensor's subnets.
— Mark Jeffrey (@markjeffrey) May 23, 2026
"This is in Alpha only. But a $TAO-powered ChatGPT-like interface. https://t.co/kASvuxTBlj .
Only using bittensor subnets (like @chutes_ai) for text and image inference.
Will be 1/250th the price of chatgpt." https://t.co/IiV6dQh2VT pic.twitter.com/SCw3eO4Aqx
Privacy protections are built into the inference layer rather than applied at the product level. DeepSeek V3.2 on Chutes runs inside a Trusted Execution Environment (TEE), a hardware-level isolation mechanism that prevents the inference provider from reading query contents during processing.
Chutes (SN64) is currently one of the most active subnets on the Bittensor network, having been the first subnet to cross a $100 million market cap following the dTAO upgrade in February 2025, doing so in nine weeks. Rayon Labs, the team behind Chutes, auto-stakes platform revenue directly into SN64 alpha token buybacks, tying inference demand to token economics.
BlueTAO is live at bluetao.ai. The team has indicated that additional features and changes are planned as the alpha progresses.
Disclaimer: This article is for informational purposes only and does not constitute financial, investment, or trading advice. The information provided should not be interpreted as an endorsement of any digital asset, security, or investment strategy. Readers should conduct their own research and consult with a licensed financial professional before making any investment decisions. The publisher and its contributors are not responsible for any losses that may arise from reliance on the information presented.