ModelCards
← All benchmarks

WER, hard prompts

We haven't written a description for this benchmark yet; see the card notes below for how each maker ran it.

Your workloadRAG over docs · 12K in · 600 out · 50K requests/mo · 30% cached
  1. 1Deepgram Flux TTS · Deepgram
    3.4%

“Best value”: no other model here reports a higher score at a lower cost for your workload. Scores are self-reported by each maker with different settings, so treat close calls as ties.

How each maker reported it

  • Deepgram Flux TTS launch post

    Deepgram's own read-aloud WER test; the post gives competitor results only as "roughly half of ElevenLabs and a third of Cartesia" and "beats the next-best model by 47%" on hard prompts, without naming exact models or values.

Only each model's own maker's number is shown; scores a maker reports for rival models are left out.