Models / OpenAI/ Whisper Large v3 Turbo

Whisper Large v3 Turbo

OpenAI · released Oct 1, 2024 · openai/whisper-large-v3-turbo

Speech to textTranscribes a recording into words

Input: audio. Output: text.InputOutput
Type
Open weightsMIT License
Languages
99
Size
0.8B

Context measured in tokens

Our take

Written Aug 3, 2026

Whisper Large v3 Turbo is a compact downloadable speech-to-text model from OpenAI that turns audio into written words at exceptional speed. It carries a permissive MIT licence and handles 99 claimed languages, though every accuracy figure we hold is English-only.

Who should pick it

Pick this for high-throughput transcription where speed beats perfection — it processes an hour of audio in under five seconds. Use it for clean recordings like read-aloud text or financial calls, with under three words wrong per hundred. Skip it if your audio is messy: meetings, podcasts or accented speech push errors past one in ten words, or if you need verified non-English accuracy.

The case for it

  • Extremely fast transcription: roughly an hour of audio in about four and a half seconds on benchmark hardware.
  • Very low error rate on clean, structured audio: about two words wrong per hundred on read-aloud speech, under three on financial calls.
  • Truly permissive MIT licence allows commercial use, modification and redistribution.
  • Broad claimed language coverage: 99 languages listed, including major world languages.

The case against it

  • Accuracy collapses on challenging real-world audio: recorded meetings see more than six times the error rate of clean read speech.
  • Even slightly harder read-aloud audio degrades sharply — a 74 percent relative increase in errors against the cleanest recordings.
  • Every accuracy figure here is English; nothing we hold measures the other 98 claimed languages.
00

How good is it?

TranscriptionTurning speech into text2 of 5Open ASR WER · 61st of 74

Words it gets right

93%

Misses roughly one word in 14, averaged over nine English test sets.

How fast it listens

783×29th of 62

an hour of audio in 5 seconds, on the board's own hardware. Your machine will differ.

Languages

99

Listed on the model card. The accuracy above is English only.

Where it struggles
Read aloudaudiobooks, clean recording2.1%
Podcasts and videoeveryday internet audio8.5%
Accented speechspeakers from many countries11.1%
Meetingsa room, several people, far microphone13.9%

Percentage of words wrong on each set, lower better. Bars are scaled to this model's own worst case, not to the board.

Which languages

English · Chinese · German · Spanish · Russian · Korean · French · Japanese · Portuguese · Turkish · Polish · Catalan · Dutch · Arabic · Swedish · Italian · Indonesian · Hindi · Finnish · Vietnamese · Hebrew · Ukrainian · Greek · Malay · Czech · Romanian · Danish · Hungarian · Tamil · Norwegian · Thai · Urdu · Croatian · Bulgarian · Lithuanian · Latin · Māori · Malayalam · Welsh · Slovak · Telugu · Persian · Latvian · Bangla · Serbian · Azerbaijani · Slovenian · Kannada · Estonian · Macedonian · Breton · Basque · Icelandic · Armenian · Nepali · Mongolian · Bosnian · Kazakh · Albanian · Swahili · Galician · Marathi · Punjabi · Sinhala · Khmer · Shona · Yoruba · Somali · Afrikaans · Occitan · Georgian · Belarusian · Tajik · Sindhi · Gujarati · Amharic · Yiddish · Lao · Uzbek · Faroese · Haitian Creole · Pashto · Turkmen · Norwegian Nynorsk · Maltese · Sanskrit · Luxembourgish · Burmese · bo · Filipino · Malagasy · Assamese · Tatar · Hawaiian · Lingala · Hausa · ba · Javanese · Sundanese

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.

Also scored, on boards we give no mark for
Accented speech 41st of 74Podcasts and video 44th of 74Financial calls 47th of 74Harder read speech 47th of 74Recorded meetings 62nd of 74Clean read speech 66th of 74European-accented speech 72nd of 74

These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done, which is why they get no rating.

Every published score for this model9 scoresEvery figure we hold, from 9 boards, with who ran it and a link to the source — including the boards no rating above is built on.
782.6independentsource ↗
7independentsource ↗
11.1independentsource ↗
2.8independentsource ↗
13.9independentsource ↗
8.5independentsource ↗
2.1independentsource ↗
3.7independentsource ↗
01

Can you run it yourself?

Fits in memory
weights load entirely on the card
Spills to system RAM
some weights offload; much slower
Too large
will not load even with offload
est
size is calculated; the verdict could change by 10%

Comfortable fit

A card many people ownFits in memory

GeForce RTX 4090 · 24 GB

Weights at Q4_K_M0.5 / 24 GBest
Spare memory21.1 GB spare
Usable context262Kwhat the spare memory holds; no published limit on record
Decode speed1663 tok/sest

Room to spare. 21.1 GB spare means a 10% error in the size would not change the answer.

One step upFits in memory

Apple M1 Pro (16-core GPU) · 32 GB

Weights at Q4_K_M0.5 / 32 GBest
Spare memory22.3 GB spare
Usable context262Kwhat the spare memory holds; no published limit on record
Decode speed289 tok/sest

Room to spare. 22.3 GB spare means a 10% error in the size would not change the answer.

On a MacFits in memory

Apple M2 (8-core GPU, 8GB unified) · 8 GB

Weights at Q4_K_M0.5 / 8 GBest
Spare memory4.3 GB spare
Usable context131Kwhat the spare memory holds; no published limit on record
Decode speed144 tok/sest

Room to spare. 4.3 GB spare means a 10% error in the size would not change the answer.

Your hardware
Checking your profile…

Memory use by level

Against a 24 GB card.

Q4_K_M
recommended
0.5 GBest
Fits in memory
Q5_K_M
0.6 GBest
Fits in memory
Q8_0
0.9 GBest
Fits in memory

All 3 sizes here are calculated, not measured. We hold no measured file for this model, so each size comes from the parameter count and every verdict above inherits that uncertainty.

Check against your own machine → · All 71 devices, with every size →

02

Or rent it from someone else

Cheapest published offer

Cheapest of 2 live listings. Picked at the widest standard context we hold, within one quantisation slice, so the numbers beside it are a price one host actually charges.

per minute of audio
$0.001
Context served
Throughput
Not measured
Current provider offers with price, context and prompt-privacy answers
ProviderPrice per minute of audioContextThroughputTrains on promptsLogs promptsZero retention
Groq$0.001not reportednot measuredUnknownUnknownUnknown
IBM watsonx$0.006not reportednot measuredUnknownUnknownUnknown

Across the 2 listings we hold: 0 say they do not train on prompts, 0 say they do and 2 do not say. 0 appear in the zero-retention registry we check; the rest are unknown to us rather than confirmed either way.

03

When we formed this view

Dates behind this page

Aug 2, 2026BenchmarkScored 11.1 on Accented speechleaderboard
Aug 2, 2026BenchmarkScored 2.8 on Financial callsleaderboard
Aug 2, 2026BenchmarkScored 13.9 on Recorded meetingsleaderboard
Aug 2, 2026BenchmarkScored 7 on European-accented speechleaderboard
Aug 2, 2026BenchmarkScored 8.5 on Podcasts and videoleaderboard
Aug 2, 2026BenchmarkScored 2.1 on Clean read speechleaderboard
Aug 2, 2026BenchmarkScored 3.7 on Harder read speechleaderboard
Aug 1, 2026ListedListed on LLMapfirst indexed by our pipeline
Aug 1, 2026BenchmarkScored 782.6 on Open ASR RTFxleaderboard
Aug 1, 2026BenchmarkScored 7 on Open ASR WERleaderboard

Prices last checked 6h ago

What we do not know about this model yet

  • We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
  • 2 of 2 listings publish no parameter list, so what their API accepts is unknown to us.
  • Nothing we hold says whether an endpoint streams, so we do not show it either way.
  • 2 of 2 listings do not say whether they train on prompts.
  • We hold no cached-input rate for any of its listings.
04

Licence and identifiers

What the licence allowsMIT License, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.

Licence

MIT License

permissiveCommercial use allowed

Fully permissive: do anything with attribution. No patent grant, unlike Apache-2.0.

Identifiers

Architecture
Dense
Modality record
audio->text
Catalogue slug
openai-whisper-large-v3-turbo

Machine-readable model card (omc.json) →

Something wrong on this page? Tell us