GPT Audio
OpenAI · released Jan 19, 2026
Speech to speechTakes speech and answers in speech
- Type
- Closed
- Input
- $2.50
- Output
- $10.00
- Cached
- None held
List price · per 1M tokens · OpenAI at 128K context · machine-readable source ↗
Our take
Written Sep 30, 2026GPT Audio is a hosted speech-to-speech model: audio goes in and audio comes back, so a spoken exchange does not have to be turned into writing first. We list no download for it, and nothing here measures how accurately it transcribes or how fast it runs, so treat it as a candidate to trial rather than a proven pick.
Reach for it when you want a spoken answer rather than a transcript, through a host. Its request capacity is large enough that long recordings need not be split up first, though reliable handling across all of it is unverified in our data. Skip it if you need measured transcription accuracy, a published language list, or a measured serving speed, none of which we hold.
The case for it
- Both listed hosts charge the same rate, so the choice between them is not a price decision.
- Audio goes in and audio comes back, so a spoken exchange does not have to be turned into writing first.
The case against it
- Word error rate is unrecorded on clean read-aloud recordings, podcasts and video, accented speech and meeting recordings, so accuracy needs a trial on recordings you can check yourself.
- Neither offer carries a measured speed, so how quickly a long recording is handled is unverified in our data.
- No language list is held, and every accuracy figure we do hold is English only, so use in another language is untested here.
How good is it?
We hold no score for this model.
The Open ASR Leaderboard is where a model that turns speech into text gets measured, and GPT Audio is not on it — so we hold no accuracy figure for it, rather than a poor one.
Where to rent it
Prices checked 2 hours ago — each listing carries its own date.
OpenAI, direct
The lab is also the cheapest we hold. The strip above and this offer are the same one, so nothing on this page undercuts OpenAI on 128K of context.
- per 1M tokens
- $2.50 in / $10.00 out
- Context served
- 128K
- Throughput
- ~80 tok/s
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouterOpenRouter's own listing | $2.50 / $10.00checked 2 hours ago | 128K | not measured | Unknown | Unknown | Unknown |
| OpenAIDirect | $2.50 / $10.00checked 2 hours ago | 128K16K max reply | 80 tok/s | No | Yesunknown period | Unknown |
OpenRouterAcross the 2 listings we hold: 1 says it does not train on prompts, 0 say they do and 1 does not say. 0 appear in the zero-retention registry we check; the rest are unknown to us.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenRouterOpenRouter's own listing | ✓ | ✓ | ✓ |
| OpenAIDirect | ✓ | ✓ | ✓ |
Tool calling: 2 of 2 listings say yes. JSON output: 2 of 2 listings say yes. Strict schema: 2 of 2 listings say yes.
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
- No independent board has scored it, so we hold no quality figures at all.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 2 listings does not say whether it trains on prompts.
- We hold no cached-input rate for any of its listings.
- We hold no batch or off-peak rate for any of its listings.
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 record for this model, and no record of published weights either — so we can neither summarise its terms nor point you at the weights.
Identifiers
- Takes in, gives back
- Text and audio in, text and audio out
- Catalogue slug
- openai-gpt-audio