Performances
RD Popularity Index
RD Growth Index
| 1h | 24h | 7d | 30d | 60d | 90d |
|---|---|---|---|---|---|
| 0.47% | 4.93% | 9.89% | 5.97% | 69.31% | -- |
Comparison
Team
Follow Updates
About c0mpute
c0mpute is a decentralized AI inference network where users pay with USDC credits and GPU providers earn USDC for processing tasks. Its core product is c0mpute code, a terminal-based coding agent that runs models across independent GPUs while keeping file edits and shell commands local. The platform offers Pro tier (browser-based, uncensored Qwen3 8B) and Max tier (local GPU, Qwen3.5 27B or SuperGemma4 26B), plus image generation via ComfyUI. The ZERO token is not used for payments but for network sustainability: half of revenue buys back and burns ZERO, half pays stakers in USDC. Stakers also get daily USDC dividends and free credits; workers staking over 500,000 ZERO receive 80% revenue share.
Narrative and differentiation: uncensored models, privacy (code stays local, secrets stripped before sending), no single provider to rate-limit or ban, open source, and worker anonymity. Market pain points addressed: centralized AI censorship, privacy leaks, single points of failure, and vendor lock-in.
Development in the past six months (2026-03-02 to 2026-09-02): launched c0mpute code agent; introduced Pro/Max service tiers with browser WebGPU and local GPU execution; implemented ZERO token buyback-and-burn and staking rewards; added ComfyUI-based image workers. Negative aspects within two years: several features remain unimplemented — real DID generation, libp2p networking, and plugin marketplace are still in progress; the network is not yet fully decentralized as core components depend on upstream projects.
Founder: leyten (Twitter: @leyten). Electrical engineer by trade, building the decentralized AI inference network @c0mputeAI. He created both the c0mpute network and its serving engine Shard (pipeline-parallel LLM inference across distributed GPUs). His prior work includes the Shard engine and the c0mpute protocol on Solana, with a focus on uncensored, private, and decentralized inference.