Phonon
A tiny text-to-speech model that runs on your device. It's for people making images, video, or audio. It's free. We haven't tried it yet.
Get notified when this tool is tested
No logged test yet. This joins the same daily newsletter, and it does not assign a tier.
No affiliate link on file. Outbound goes to huggingface.co. Identity checked October 22, 2026. Scores were not part of that check.
Grade
We have not tried it yet. The letter on the list is our best guess.
- Grade
- B · Best guess
- Score
- Test pending
- Price
- Free
Best for
Jobs people use this for.
Test details
What's new · October 19, 2026Gradium published Phonon, on-device text-to-speech. The 31M English model card lists 31 million parameters, a 64.5 MB q8 file, and Apache 2.0. Larger 40M weights on that card are CC BY-NC. Agent test pending.Source
Gradium's on-device text-to-speech. The Phonon 31M English Hugging Face card lists 31 million parameters, a 64.5 MB q8 weights file, and an Apache-2.0 license. The same card lists larger 40M models under CC BY-NC. We have not run it.
Test pending. Open weights on Hugging Face. We have not run Phonon.
- Status
- Not run
- Date
- —
- Time to first output
- —
- Cost
- —
- Result
- —
- Evidence
- None
Queued task. Render a fixed script in a chosen voice. Log time to the first usable take and artifacts.
FAQ
What is the agent-tested verdict for Phonon?
A tiny text-to-speech model that runs on your device. It's for people making images, video, or audio. It's free. We haven't tried it yet. We have not tried it yet. The letter on the list is our best guess.
How much does Phonon cost?
Free. Phonon 31M English weights are Apache-2.0. The 40M weights on the same card are CC BY-NC.
Do sponsors outrank this tool?
No. A sponsored slot is labeled and sold separately from tiers. It cannot move this listing.