Skip to content
Hot storyPast story

Inworld shares Cekura TTS benchmark, says TTS-2 leads in latency and accuracy

1 report1 sourceUpdated 2026-10-22

What happened

AI overview

Inworld AI shared a text-to-speech (TTS) benchmark published by third party Cekura on X, saying its TTS-2 series leads in latency and accuracy. According to the test data, TTS-2 Flash has a median latency of 46ms from receiving a complete sentence to starting speech. TTS-2's estimated read accuracy is 93.4%, and TTS-2 Flash's is 93.2%. The test covered 26 models and 30,564 requests, using 104 sentences. Each model read the sentences three times in the same cloud region. Accuracy was estimated by an automatic speech-to-text scorer, and the content included phone numbers, dates, code, and medical terms. Inworld thanked Cekura for publishing its methodology and audio samples so others could compare the results for themselves.

AI-written from the coverage · updated 2026-10-22

Coverage timeline

Follow the coverage to see the story from different sides.

Oct 22
  1. Inworld AI
    Cekura releases TTS benchmark, Inworld's TTS-2 series leads in latency and accuracy

    Cekura's text-to-speech benchmark covers 26 models and 30,564 requests. Inworld's TTS-2 Flash has a median latency of 46ms from receiving a complete sentence to starting speech, and TTS-2's estimated read accuracy is 93.4%. The test used 104 sentences, with each model reading them three times in the same cloud region, and accuracy was estimated by an automatic speech-to-text scorer. TTS-2 Flash's estimated accuracy is 93.2%, and the test content included phone numbers, dates, code, and medical terms.