Solaria 3
Gladia
Speech to textTranscribes a recording into words
- Type
- Closed
- Input
- None held
- Output
- None held
- Cached
- None held
We don't hold a list price for this model yet · hosted only — no weights published
Our take
Written Sep 4, 2026Solaria 3 is a proprietary speech-to-text model from Gladia that turns audio into written words. It scores above average on every English audio condition we track, from clean read-aloud speech to noisy meeting recordings, though it is never the absolute best on any single one.
Pick this for general-purpose English transcription where consistent above-average accuracy matters more than winning on one condition. It is a strong choice for podcasts, video and meeting-room audio, and for accented speech where it still beats most rivals. Skip it if you need hosted access today — we list no current offers — or if you want the lowest possible error rate on a specific audio type.
The case for it
- Above-average accuracy on every English audio condition we measure, from read-aloud to meetings.
- Near the top of the field on clean professional audio: financial calls and parliamentary speech.
- Five languages listed, broader than many proprietary API transcription models.
The case against it
- No hosted access currently available; proprietary weights with no published parameter count or release date.
- Never the absolute top rank: even its best condition is 0.4 points above the field leader.
- Accented speech is its weakest showing at 10.5% error, three points above the leader on that condition.
How good is it?
A speech-to-text model for turning recordings of talks, interviews and accented speakers into written text.
- turning spoken English into written textOpen ASR WER · 19th of 76
- transcribing speakers with a range of accentsAccented speech · 11th of 76
- transcribing podcasts and video audioPodcasts and video · 10th of 92
TranscriptionTurning speech into text4 of 5Open ASR WER · 19th of 76
95.4%
Misses roughly one word in 22, averaged over nine English test sets.
5
Stated by the leaderboard; we do not hold the list itself.
Percentage of words wrong on each set, lower better. Bars are scaled to this model's own worst case; the placing beneath each rate is against every model measured on that set.
The figures above come from the Open ASR Leaderboard, an independent public test that runs every model on the same recordings. It is the only measurement of transcription quality we know of, so there are no other scores to show.
Each of these is the same transcription job on a different kind of recording, so together they say where it holds up and where it slips — not how closely it follows an instruction.
Every published score for this model8 scoresEvery figure we hold, from 8 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Where to get it
We hold no priced listing for Solaria 3.
There is no copy to download and no host in our price data, so Gladia is where to look. We watch OpenRouter, the provider APIs we track and the LiteLLM price set; this version appears in none of them, which is a gap in what we collect rather than a statement about what Gladia sells.
When we formed this view
Recent changes
What moved
first indexed by our pipelineEach date is the day we first saw the change, or the day the maker announced it.
What we do not know about this model yet
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- We don't hold a list price for this model yet — the gap is ours, not the lab's.
Licence and identifiers
What the licence allowsWe hold no licence record for this model. Inside are the identifiers you need to pull it — its Hugging Face repo where we have one, our slug and a machine-readable card.
Licence
We hold no licence row of this model's own. A source states its weights are not published, so the determination that governs it is the one for closed weights, API access only.
Identifiers
- Takes in, gives back
- Audio in, text out
- Catalogue slug
- gladia-solaria-3