Speechmatics

Melia 1 is Speechmatics' multilingual model, launched June 2026, transcribing speech that switches among 55+ languages in one pass and labelling each word's language. It requires language set to multi and lacks custom dictionary, confidence scores and translation.

Enhanced is the highest-accuracy model in Speechmatics' batch API, from the Ursa 2 generation (October 2024), covering 55+ languages. Its options include diarization, translation and a medical-domain variant, with slower turnaround than Standard.

Speechmatics' top-accuracy model for live audio, running the Enhanced tier over its realtime WebSocket API. Speechmatics says its Ursa 2 generation brought realtime latency under one second; speaker diarization and partial transcripts are supported across 55+ languages.

Standard is Speechmatics' default batch model, feature-identical to Enhanced but tuned for throughput and turnaround rather than peak accuracy. Speechmatics recommends it for archival transcription, content indexing and large-scale captioning across 55+ languages.

Realtime counterpart of Speechmatics' Standard tier, giving up some accuracy against Enhanced in exchange for speed on live audio. It keeps the full Standard feature set, including speaker diarization, custom dictionary and confidence scores.