NEAR AI Cloud adds OpenAI-compatible API with private inference

  • NEAR AI launched a cloud API compatible with OpenAI, enabling developers to retain existing clients while changing the base URL or endpoint and authentication key.
  • NEAR says TEE-hosted models, including GLM 5.2 variants, use Intel TDX and NVIDIA confidential computing hardware with cryptographic attestation for each request.
  • The newer report dates the announcement to Sept. 11, while the older account dates the full launch to Dec. 3, 2025; later developments included staking-based AI payments and Virtual $NEAR Day programming.

NEAR AI, the artificial intelligence unit of NEAR, launched a cloud API compatible with OpenAI, allowing developers to redirect existing clients by changing the base URL or API endpoint and authentication key. The service supports chat and legacy completions, Python and JavaScript SDKs, tool calling and structured outputs, with requests processed through gateway or model-specific endpoints in Trusted Execution Environments (TEEs). NEAR says models including GLM 5.2 variants use Intel TDX and NVIDIA confidential computing hardware and provide cryptographic attestation for each request. The newer account dates an announcement to Sept. 11, while the older account dates the full compatibility launch to Dec. 3, 2025. Around July 2026, NEAR added staking-based payments for AI compute credits, and its model catalog expanded between July and September 2026 with LiteLLM, Aider and Continue integrations. Virtual $NEAR Day covers formal verification, artificial intelligence, quantum technology, privacy, AI staking and cross-chain capabilities. At the time of the newer report, NEAR traded at $2.6, up 9.04%, according to CoinMarketCap.

The information on this website is generated using AI and we cannot guarantee its accuracy. Please use it as reference information only.