The vercel/ai repository added a provider documentation page for NEAR AI Cloud, an OpenAI-compatible TEE-backed inference service, including setup instructions, code examples for text generation and streaming, and an entry in the OpenAI-compatible providers index.
Adds official documentation and setup guide enabling AI SDK developers to integrate NEAR AI Cloud as an OpenAI-compatible inference backend without guessing configuration
Developers using the Vercel AI SDK can now discover and integrate NEAR AI Cloud's TEE-backed inference endpoint without searching external sources or trial-and-error configuration, reducing integration friction for privacy-sensitive inference workloads. The docs cover a request-body transform to avoid forwarding `reasoning_effort`, which may cause unexpected behavior if omitted by developers not reading carefully. Watch for whether the documented endpoint URL and key naming remain stable as NEAR AI Cloud matures, and whether the reasoning_effort transform becomes a default SDK behavior.