Chutes SN64Price $ 24.76

Chutes.ai is a decentralized, serverless AI compute platform built as Subnet 64 on the Bittensor network, providing on-demand inference and scaling for open-source AI models. It enables deployment and running of state-of-the-art LLMs, image/video/speech models, embedding models, and moderation tools with fast access to new releases (often minutes after launch) and support for custom models or user code.

The platform uses a permissionless miner network (miners earn TAO from Bittensor emissions by processing prompts/inference jobs) combined with Trusted Execution Environments (TEE) for secure, private compute where even GPU providers cannot access user data.

Key mechanics include OpenAI-compatible APIs for easy integration, usage-based pricing per million tokens (input/output), permanently hot models for low latency, and features like batch processing, long-running jobs, and model training support.

It powers consumer apps such as Chutes Chat for general interaction, Chutes Search (AI-powered search tool), and Fictio for custom experiences. The platform emphasizes defense-in-depth security in an adversarial environment with anonymous miners, verifiable integrity, and privacy via TEEs.