Capabilities
What it does
Five modalities, one key
LLM, decision, image, audio and embedding models answer at api.sylphx.com/v1 with one Sylphx key, and usage from all of them lands on one bill.
Priced in each model’s own unit
Tokens for language and embeddings, characters for speech, images or megapixels for pictures, seconds of audio for realtime transcription. Every response reports what it used.
Your data policy, per model
Requests take the private default: no training, and zero retention where the model offers it. A route that fails is skipped and retried later, with no change on your side.
Links