Support for OpenAI Embedding, Speech-to-Text, and Text-to-Speech models.

[ember lotus]

Could you please add support for the following embedding, Speech-to-Text, and Text-to-Speech models?

For embedding models, I need: text-embedding-ada-002, text-embedding-3-large, and text-embedding-3-small. For Speech-to-Text, I need the whisper-1 model. For Text-to-Speech, I need support for tts-1 and tts-1-hd models.

[covert inlet]

@hexed heath @fluid bramble @covert token your stance on (whether/when to provide) those widely-asked services?

[fluid bramble]

Once we get our infra in a decent spot where a massive surge in traffic doesn't take us down for a whole hour (or more), we will start adding these

[covert inlet]

understood, fingers crossed!+)

[cerulean wolf]

Is it time? Had a great experience with Openrouter. It'd be great if we can use embedding and rerank models on the platform.

[hexed heath]

cc @slate glade

[slate glade]

@cerulean wolf what do you mean by rerank models?

[hexed heath]

@slate glade example: https://cohere.com/rerank

[zenith oxide]

@slate glade - there's lots of "AI" related modalities vendors.
chat (openrouter supports), completion (code auto-complete in IDEs, for example) embedding (e.g for RAG or searching by ideas, instead of exact words), image generation, Text to Speech, Speech to Text. Image to video. text to songs. Re-ranking (RAG/embedding related)

See https://countless.dev/ for some of them.

[eternal mesa]

Is there any example showing how to use embedding models through OpenRouter?

[lyric hatch]

there's no embedding models in OpenRouter