TEE (Trusted Execution Environment) is a hardware-level security enclave that isolates your code and data. On ComputeFlux, all API requests, model inferences, and token forwarding run inside TEE, meaning even the platform operator cannot see your data. This provides stronger privacy guarantees than traditional encrypted proxies.

Every request is encrypted end-to-end and processed inside TEE secure enclaves. Your API keys, prompts, and request logs are never stored or visible to anyone — including ComputeFlux. Combined with blockchain-based settlement, you get full transparency on payments with zero data exposure.

No. ComputeFlux is 100% Web3 native. Simply connect your wallet (MetaMask, Polkadot.js, WalletConnect, etc.) and start using the service immediately. No email, no ID verification, no credit card.

Connect your Web3 wallet, create an API key in the dashboard, and use our unified endpoint to access models across 10+ blockchains. Check our docs for quickstart guides with OpenAI-compatible SDKs, LangChain, LlamaIndex, and more.

Deploy a node using Docker on any device with idle compute (GPU, NAS, router, or bandwidth). Your node serves API requests and earns real crypto tokens — not points. Withdraw instantly to your preferred low-gas chain like Solana, Base, or Polygon.

ComputeFlux supports 10+ blockchains including Ethereum, Solana, Base, Optimism, Arbitrum, Polygon, Sui, Aptos, Polkadot, and more. Both API routing and payment settlement work across all supported chains.

Unlike OpenRouter, Infura, and Alchemy which are centralized and may require KYC, ComputeFlux is fully decentralized with TEE privacy protection. Your requests are invisible to the platform, you pay with crypto (no credit card), and you get multi-chain access through a single API. It's the privacy-first alternative for Web3 developers.

We think the core difference is not any single technical point — TEE, price, or the number of models — but combining these capabilities into a product that ordinary users and developers can actually use. In short: • One ComputeFlux API / account gives you access to the AI capabilities of the entire network. • The underlying compute can come from different GPU providers, models, and infrastructure — users don't need to understand that complexity. Going forward, we will add: • Unified access to multiple models and providers • Automatic routing by price, speed, and stability • Provider fallback for better availability • TEE / privacy-preserving inference • Open GPU providers, letting third-party compute join the network • Agents and other AI capabilities So TEE will be an important part of ComputeFlux, but we don't define ComputeFlux as merely a TEE inference network. What we really want to build is an open AI compute marketplace and a unified AI infrastructure layer.

We don't think Web3 or decentralization by itself should be the reason to use ComputeFlux. What users ultimately care about is: can I get more AI capabilities, more reliably, in a simpler way and at a more competitive price? Today most of the compute on ComputeFlux is still provided by us. Our pricing is benchmarked against the market API prices of the corresponding models, and we aim to stay as competitive as possible. As more GPU providers join, the long-term pricing mechanism will gradually become: More providers → providers compete for workloads → higher GPU utilization → lower inference cost → more users. That is also why we don't want to promise a fixed “30% or 50% cheaper than OpenAI/OpenRouter” for now — it varies greatly by model, hardware, and market conditions. But we do want price competition to become part of the network itself through the marketplace. ComputeFlux already has incentive mechanisms such as Points. As the provider network opens up, GPU owners can go from being pure AI consumers to compute providers: Idle GPU → ComputeFlux provider → real AI workloads → revenue. This forms a natural economic loop: More AI users → more real compute demand → providers earn more → more GPUs join → stronger compute competition → users get better prices and service.

Yes. Your API key is stored inside the TEE secure enclave and is never exposed to the platform or any third party. Even if a node is compromised, the TEE hardware protection ensures your credentials remain encrypted and inaccessible.