Amazon models, access and benchmark guide
Amazon model and access guide with a practical US-English evaluation brief and links to publisher materials.

👁 111 reads
An AWS quote for Nova speech needs a region, model ID and input/output units. An older endpoint can remain visible in a rate sheet after the window for new use closes. I would match the regional deployment list and billing table before opening the microphone.
Choosing the first job
In us-east-1, confirm the selected Nova route and measure input speech, output speech and text separately on a dummy call.
Models in the November 17 snapshot
| Model and track | Placement | Reported score |
|---|---|---|
| Nova 2.0 Sonic (Mar 2026) (AA live voice) | 16 | 964 Elo |
| Nova Sonic (AA live voice) | 22 | 907 Elo |
Release record. Publisher documentation identifies the product family. The ranked configurations include Nova 2.0 Sonic (Mar 2026), Nova Sonic.
Google · Your sourcesQuantslant higher in your Google - one clickAdd to preferred sources →Sources
- https://aws.amazon.com/nova/ (accessed November 17, 2026)
- https://docs.aws.amazon.com/nova/latest/userguide/speech.html (accessed November 17, 2026)
- https://artificialanalysis.ai/speech-to-speech/arena (accessed November 17, 2026)
More on this from QUANTSLANT
ai tools·AI image editors ranked on changing one detailSee how the image-editing leaderboard changes when a generator must keep the rest of a source photo intact.