Speech translation models

Live speech translation models: speech in, translated speech or text out, as the speaker talks. Text-only translation models are language models. Price per minute of audio on the provider’s own API where it prices by audio length; per 1M audio tokens where it prices by tokens. Rows show standard prices; other price tiers are on each model’s page. Dots show where each value came from; a hollow dot means it is not verified yet.

11 of 11 speech translation models, 3 dated snapshots under their models

11 of 11 models expanded