Add OrcaRouter as an optional AI provider

Hi maintainers,

We'd like to propose adding [OrcaRouter](https://www.orcarouter.ai) as an optional AI provider alongside the providers already supported by the project.

OrcaRouter exposes an OpenAI-compatible API and uses standard API-key authentication, so the integration should be able to reuse the project's existing OpenAI/provider abstraction with minimal changes.

## What this would add

- Access to multiple chat, reasoning, image, and video models through one endpoint
- Automatic model routing and provider failover
- Standard API-key authentication
- Prompt caching
- Usage tracking, budgets, and team access controls

The integration would not replace or modify any existing provider. OrcaRouter would simply appear as another provider option that users can configure with their own API key.

## Existing OSS ecosystem

OrcaRouter is already integrated with or available across a growing open-source ecosystem, including RAGFlow, Dify, goose, promptfoo, NocoBase, DB-GPT, CAMEL, and models.dev / OpenCode.

## Partner program disclosure

OrcaRouter operates an optional open-source partner program. Approved OSS projects can receive a 5% revenue share from OrcaRouter usage attributed to their integration.

Participation in the partner program is not required for the provider integration, and we're happy to follow any disclosure or governance requirements maintained by the project.

More information:

https://www.orcarouter.ai/built-with

If this sounds useful, we'd be happy to discuss the integration details and, with the maintainers' approval, open a PR with the implementation.

I'm an engineer on the OrcaRouter team.

The issue names no files or tests. Start by locating the existing OpenAI/provider abstraction and configuration entry points, then verify OrcaRouter's API and authentication requirements. Done means OrcaRouter is available as an optional provider without changing existing providers, with the integration covered by the project's tests.

Written by the indexing model from the issue text.