Gemini /v1beta/voices vs finding a voice id on Sume

On Sume you find a voice by listing avatars, not by calling a voices route. The TTS contract says to list GET /v1/avatar-1.0/avatars and use any avatar whose voice.status is ready; Sume then resolves that avatar's TTS voice when you submit. Gemini's new /v1beta/voices is Google's own listing and its ids are not Sume ids.

What did Gemini ship?

The Gemini API changelog entry dated September 22, 2026 (read 2026-09-30) says Gemini 3.8 Flash TTS and Flash-Lite TTS are generally available, along with the Gemini API Voices endpoint at /v1beta/voices. It describes the endpoint as letting you query 150+ prebuilt and custom voices, with consent verification for custom ones.

How do I get a voice on Sume?

The two products solve voice discovery in different places.

What does the request look like?

Send a transcript and the avatar reference. The schema notes that avatar_id and avatar_handle are request fields, not properties of voice, and that if both an avatar and voice.id are set they must match or the request fails with 400. A caller holding only the contract never needs a voice id it cannot obtain.

curl https://api.sume.com/v1/avatar-1.0/avatars \
  -H "Authorization: Bearer $SUME_API_KEY"

curl -X POST https://api.sume.com/v1/tts-1.0/generate \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: voice-pick-001" \
  -d '{"transcript": "Hello from Sume.", "avatar_handle": "acme"}'

Can I reuse a Gemini voice name on Sume?

No. A voice name from another TTS ecosystem is rejected with a 400; see the invalid voice id post. If you were weighing Gemini TTS itself, Gemini TTS API: which engine Sume routes covers what Sume's Router lists.