Context

prompts.chat has come a long way from its Awesome ChatGPT Prompts roots, and the project now reads like a platform, not a list, for anyone who works with AI prompts. Two things I value are the AI-powered semantic search that translates non-English queries to English before embedding, and the ability to self-host a branded, private prompt library for an organization.

Both features draw on the same config surface: the AI layer in src/lib/ai reads OPENAI_API_KEY, points the OpenAI SDK at an optional OPENAI_BASE_URL, and picks models via env vars like OPENAI_GENERATIVE_MODEL. For someone self-hosting prompts.chat, an OrcaRouter endpoint behind that same base-URL setting would let generation, translation, and semantic-search calls route across several upstream providers with automatic failover, instead of depending on a single provider key.

Proposal: OrcaRouter as an optional provider

I'm an engineer on the OrcaRouter team. I'd like to propose adding OrcaRouter as an optional provider for prompts.chat's self-hosted AI features. This would not replace or change any existing provider — it would simply be one more choice behind the OPENAI_BASE_URL seam your code already supports, so existing deployments and API keys are unaffected.

OrcaRouter exposes an OpenAI-compatible API and uses standard API-key authentication. Capabilities most relevant to prompts.chat users:

• Automatic model routing and provider failover — requests for generation, translation, or semantic search go to the best available upstream model and retry on another provider when one is rate-limited or down. For orgs running prompt-builder and search in production, that removes a single point of failure.
• Usage tracking and budgets — one dashboard across all routed providers, with per-key budgets, which suits admins who run a shared library and want AI spend visible and capped.

Nothing here is implemented or tested in this repo — this is a proposal to gauge interest. If maintainers are open to it, the integration point is narrow: the OpenAI SDK construction in src/lib/ai/generation.ts and src/lib/ai/embeddings.ts, which already honors OPENAI_BASE_URL and standard key auth.

OrcaRouter is already used in the open-source ecosystem by similar self-hosted AI projects, including Dify, RAGFlow, and goose.

Disclosure

Full transparency: OrcaRouter runs an optional open-source partner program in which approved OSS projects can receive a 5% revenue share from OrcaRouter usage attributed to their integration. Participation is not a prerequisite for integrating OrcaRouter, and I'm glad to follow prompts.chat's disclosure and governance expectations either way.

More on projects built with OrcaRouter: https://www.orcarouter.ai/built-with

Would the maintainers be open to OrcaRouter as an optional endpoint for self-hosted AI features? If so, I'd be happy to open a PR with the implementation once we agree on the approach.