Skip to content

v4.0.5 — kie's curated voice catalog for TTS/dialogue

Choose a tag to compare

View all tags
@elibarnett elibarnett released this 11 Jun 22:00
· 30 commits to main since this release
v4.0.5
ad80fa5
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Patch release. kie.ai only accepts a curated set of ~67 ElevenLabs voices and rejects all others with 500: This voice is not within the range of allowed options — the MCP never surfaced the list, so VO work was effectively locked to the single default voice (James).

Fixed

  • Ships kie's full voice catalog (67 voices, id + name + style), shared by turbo-2-5, multilingual-v2, and text-to-dialogue-v3. generate_tts and generate_dialogue accept a voice name ("Bella", "Viking Bjorn", "Aria") or allowlisted ID, validate client-side, and an unknown value returns the complete catalog so agents self-correct without burning a request. Duplicate names (Mark, Hope, Jessica Anne Bogart, Viraj ×2 each) disambiguate by ID. (#16, closes #15)

Verified live 2026-06-11 over MCP stdio: voice by name → valid MP3; unknown voice → catalog error with all 67 entries, no API call.

Full details in CHANGELOG.md.

Assets 2
Loading