feat(llm): add OrcaRouter as a named LLM provider

[XiaoHuo888-hue]

Summary

Add OrcaRouter as a named LLM provider for LLM-based extraction. OrcaRouter is an OpenAI-compatible gateway at https://api.orcarouter.ai/v1; model ids use the orcarouter/<model> prefix (e.g. orcarouter/auto, orcarouter/free). It also runs gateway-level, zero-trust security for AI agents on the same endpoint — screening every prompt/response and governing every tool call on a default-deny basis, with no application code changes.

The pinned unclecode-litellm build has no native orcarouter/ provider prefix, so Crawl4AI routes orcarouter/<model> through the openai provider while keeping the full model id intact and pointing base_url at the gateway.

llm_config = LLMConfig(
    provider="orcarouter/auto",
    api_token=os.getenv("ORCAROUTER_API_KEY"),
)

base_url defaults to https://api.orcarouter.ai/v1 and ORCAROUTER_API_KEY is auto-resolved, so users can just pass provider="orcarouter/auto".

List of files changed and why

• crawl4ai/config.py — register orcarouter in PROVIDER_MODELS_PREFIXES (auto-resolves ORCAROUTER_API_KEY), add ORCAROUTER_BASE_URL and orcarouter_litellm_params() helper.
• crawl4ai/utils.py — route orcarouter/ providers in perform_completion_with_backoff, aperform_completion_with_backoff, extract_blocks_batch, and get_text_embeddings.
• crawl4ai/async_configs.py — LLMConfig defaults base_url to the gateway for orcarouter/ providers.
• crawl4ai/cli.py — route orcarouter/ in the -q streaming path and mention it in the provider hint.
• docs/md_v2/extraction/llm-strategies.md — document OrcaRouter usage.
• README.md — add an OrcaRouter comment to the LLM extraction example.
• tests/test_orcarouter_provider.py — new unit tests for routing + auto key/base_url resolution.

How Has This Been Tested?

• New unit tests (tests/test_orcarouter_provider.py, 9 tests) pass.
• Live-tested against the OrcaRouter gateway through the actual Crawl4AI completion path (perform_completion_with_backoff with provider="orcarouter/auto"): HTTP 200, returns ORCA-LIVE-OK.
• Live-tested embeddings via get_text_embeddings through OrcaRouter: HTTP 200.
• python -m py_compile on all modified modules passes.

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have added/updated unit tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes

Disclosure: I'm an engineer on the OrcaRouter team.