mirror of
https://github.com/BerriAI/litellm.git
synced 2026-09-26 01:12:21 +00:00
2f996e0c10
Routes edenai/<provider>/<model> deployments to Eden AI's OpenAI-compatible API at https://api.edenai.run/v3 with EDENAI_API_KEY and EDENAI_API_BASE, and bills the per-request cost Eden reports instead of a price-map estimate, streams included Chat completions and the edenai/* wildcard (Eden's live catalog) go through the OpenAI-compatible base. /v1/responses and /v1/messages reach Eden's native endpoints so stateful Responses and untranslated Anthropic payloads work. Embeddings, transcription, speech, image generation and video generation get per-surface configs that lift Eden's cost from the body or the x-edenai-cost header. A completed Responses stream event that reports usage.cost is now billed that number for every provider, matching chat streams Admin UI provider entry, README row, endpoint matrix and provider docs tests updated; tests live under tests/test_litellm/llms/edenai |
||
|---|---|---|
| .. | ||
| litellm-dashboard | ||
| Dockerfile | ||
| nginx.conf | ||