Text-to-speech · registry
Model metadata with explicit evidence gaps.
Catalogue review: October 7, 2026. Historical models remain explicitly named. Unsupported MOS values and measured labels have been withdrawn; a source-matched listener study is required before a score is shown.
| Model | Type | Evidence tier | MOS | MOS interpretation | Source | Deployment | License | Capabilities |
|---|---|---|---|---|---|---|---|---|
Cartesia Sonic 2 | API | metadata only | Not verified | No verified listener MOS | metadata only | api | provider terms | streaming · cloning · voice design |
Dia 1.6B | open source | metadata only | Not verified | No verified listener MOS | metadata only | local | Apache-2.0 | cloning · voice design |
ElevenLabs Flash v2.5 | API | metadata only | Not verified | No verified listener MOS | metadata only | api | provider terms | streaming · cloning · voice design · code-switching |
ElevenLabs Turbo v2.5 | API | metadata only | Not verified | No verified listener MOS | metadata only | api | provider terms | streaming · cloning · voice design · code-switching |
F5-TTS | open weights | metadata only | Not verified | No verified listener MOS | metadata only | local | MIT code; CC-BY-NC pretrained weights | cloning · code-switching |
Fish Audio S2 Pro | open weights | metadata only | Not verified | No verified listener MOS | metadata only | local | fish-audio-research-license | streaming · cloning · voice design · code-switching |
Fish Speech 1.5 | open weights | metadata only | Not verified | No verified listener MOS | metadata only | local | CC-BY-NC-SA-4.0 weights; version-specific code | cloning · code-switching |
Gemini 2.5 Flash TTS | API | metadata only | Not verified | No verified listener MOS | metadata only | api | provider terms | streaming · voice design · code-switching |
Gemini 2.5 Pro TTS | API | metadata only | Not verified | No verified listener MOS | metadata only | api | provider terms | streaming · voice design · code-switching |
Google Chirp 3 HD | API | metadata only | Not verified | No verified listener MOS | metadata only | api | provider terms | streaming · cloning · voice design · code-switching |
Gradium TTS | API | metadata only | Not verified | No verified listener MOS | metadata only | api | provider terms | streaming · cloning · voice design |
Kokoro v1.0 | open source | metadata only | Not verified | No verified listener MOS | metadata only | local, edge | Apache-2.0 | streaming |
OpenAI TTS HD | API | metadata only | Not verified | No verified listener MOS | metadata only | api | provider terms | streaming |
Orpheus TTS | open source | metadata only | Not verified | No verified listener MOS | metadata only | local | Apache-2.0 | cloning · voice design |
Parler-TTS | open source | metadata only | Not verified | No verified listener MOS | metadata only | local | Apache-2.0 | voice design |
Piper | open source | metadata only | Not verified | No verified listener MOS | metadata only | local, edge | GPL-3.0 engine; voice-specific weights | streaming |
PlayHT 3.0 | API | metadata only | Not verified | No verified listener MOS | metadata only | api | provider terms | streaming · cloning · voice design |
Sesame CSM | open source | metadata only | Not verified | No verified listener MOS | metadata only | local | Apache-2.0 | voice design |
Spark-TTS | open source | metadata only | Not verified | No verified listener MOS | metadata only | local | Apache-2.0 | cloning · voice design · code-switching |
Supertonic 3 | open source | metadata only | Not verified | No verified listener MOS | metadata only | local, edge, browser | OpenRAIL-M | cloning · voice design · code-switching |
XTTS v2 | open weights | metadata only | Not verified | No verified listener MOS | metadata only | local | CPML | cloning · code-switching |