Growth · Bittensor subnet
Chutes SN64 alpha
Type Bittensor dTAO alpha token
Tokens/day 160B
Cumulative 9.1T
Users 400K
Verified revenue $1.3–2.4M
What it does
Chutes is a serverless GPU inference network on Bittensor. It deploys Llama, Qwen, and DeepSeek models at roughly 85% below AWS pricing, letting developers and apps pay for LLM inference against a decentralized GPU pool.
Fundamentals
- 160B tokens/day processed; 9.1T cumulative.
- 400K users.
- $1.3–2.4M verified external revenue — the highest verified on any Bittensor subnet.
- Real developers using it for real inference, not grant-farming.
Tokenomics
- Auto-staking buyback: external revenue directly buys SN64 alpha. Non-speculative demand tied to actual usage.
- Bittensor emissions ($52M/yr to top subnets) still dwarf ecosystem revenue, but Chutes is the subnet where the flywheel most clearly turns positive.
Catalysts (6–12mo)
- Continued Bittensor emissions weighted to top subnets.
- Migration from grant-farming subnets to real-usage models — Chutes benefits by category.
- Any expansion of enterprise inference customers.
Risks
- Emissions ($52M/yr) still dwarf revenue at the subnet-ecosystem level.
- TAO price weakness drags all subnets via denominator effect.
- Alpha tokens are illiquid vs majors — position sizing constrained by market depth.
Strong Buy. Best-fundamentals subnet in the Bittensor ecosystem with a functioning revenue → buyback loop. The right way to bet on Bittensor is to bet on the winning subnet, not TAO itself.