Skip to content

OpenAI · Transcription

Whisper Large V3

OpenAI 🇺🇸TranscriptionWhisper familyOpen weights. One of its sources says the weights are publicly available.
From Neosantara's listing:
Whisper Large v3 is OpenAI's state-of-the-art speech recognition model, providing high accuracy across diverse audio conditions and multilingual support.
Market median price
$0.0018 /min
Max context
4K
Listings
20
Providers
20

Evidence and sources

🤗 openai/whisper-large-v3 Model architecture, as its sources state it Good value (55.6). Cheaper per task than its benchmark results predict. From 1 benchmark with measured task costs. Little data behind it, so treat it as a rough estimate.How Value for money worksReleased 7 Nov 2023Maker HQ: US (according to OpenAI's own website)
All rankings: #667 on OpenRouter today #699 on OpenRouter this week #761 on OpenRouter this monthSources: DeepInfra data, Infercom data, Venice.ai data, Models.dev data, Neosantara Inference data, Hugging FacePrimary data from DeepInfra from Infercom: architecture from Venice.ai: primary mode from Models.dev: weights availability from Neosantara Inference: description from Hugging Face: release date
Tags from its sources:textaudio

Advantages and trade-offs

Compared with 10 representative transcription models

Advantages

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 04:34 UTC.

Input & output

as stated by the model's providers
Text
Audio
Text

Capabilities

Supported operations and provider-stated features

Structured outputAudio input

Listings

 Evrocopenai/whisper-large-v3Source: Models.dev data448$0.0023$0.0023—
 Near AIopenai/whisper-large-v3Sources: Models.dev data, NEAR AI Cloud data448$0.01$0.01—
 Privatemode AIwhisper-large-v3Source: Models.dev data448$0.0162$0.00—
 InfercomRegions: 🇪🇺 EUWhisper-Large-v3Source: Infercom data448—
 Together AIopenai/whisper-large-v3Sources: LiteLLM data, Together AI dataNot published by the provider$0.27$0.85$0.0015
 T Systemswhisper-large-v3Source: T-Systems LLM Hub448—
 Fastrouteropenai/whisper-large-v3Source: FastRouter data4K——$0.00045
 NanoGPTWhisper-Large-V3Source: NanoGPTNot published by the provider——$0.0005
 Groqwhisper-large-v3Sources: Groq API data, LiteLLM data, Models.dev data448——$0.0018
 Neosantarawhisper-large-v3Source: Neosantara Inference dataNot published by the provider——$0.0019
Show all 14 listings

Showing 10 of 14: the cheapest per 1M input tokens from each provider, then listings priced per minute.

Also sold by Eurouter, Scaleway, Simplismart, Venice.

6 listings without a price are hidden.

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Speech

Pashto ASR10 related benchmarks · 10 results
FLEURS ps_af CER (zero-shot)
Source: Pashto ASR paper data
table: II, dataset: FLEURS ps_af, metric: CER, protocol: zero-shot
CV24_filtered CER (zero-shot)
Source: Pashto ASR paper data
table: II, dataset: CV24_filtered, metric: CER, protocol: zero-shot
CV24_filtered WER (zero-shot)
Source: Pashto ASR paper data
table: II, dataset: CV24_filtered, metric: WER, protocol: zero-shot
FLEURS ps_af script Ar/Da/Ur (zero-shot)
Source: Pashto ASR paper data
table: III, dataset: FLEURS ps_af, metric: script Ar/Da/Ur, protocol: zero-shot
FLEURS ps_af script Em (zero-shot)
Source: Pashto ASR paper data
table: III, dataset: FLEURS ps_af, metric: script Em, protocol: zero-shot
FLEURS ps_af script In (zero-shot)
Source: Pashto ASR paper data
table: III, dataset: FLEURS ps_af, metric: script In, protocol: zero-shot
Show 4 more resultsHide 4 results
FLEURS ps_af script La (zero-shot)
Source: Pashto ASR paper data
table: III, dataset: FLEURS ps_af, metric: script La, protocol: zero-shot
FLEURS ps_af script Pashto (zero-shot)
Source: Pashto ASR paper data
table: II, dataset: FLEURS ps_af, metric: script Pashto, protocol: zero-shot
FLEURS ps_af script Ps (zero-shot)
Source: Pashto ASR paper data
table: III, dataset: FLEURS ps_af, metric: script Ps, protocol: zero-shot
FLEURS ps_af WER (zero-shot)
Source: Pashto ASR paper data
table: II, dataset: FLEURS ps_af, metric: WER, protocol: zero-shot

Other

Data credits: Hugging Face official benchmark leaderboards; licence varies by benchmark · Hanif Rahman. Benchmarking Multilingual Speech Models on Pashto: Zero-Shot ASR, Script Failure, and Cross-Domain Evaluation, arXiv:2604.04598v1. Numeric/protocol extract from Tables II, III, V and VI; CC BY 4.0 article grant verified. (licence)

Also known as: groq/whisper-large-v3, openai/whisper-large-v3, scaleway/openai/whisper-large-v3, together_ai/openai/whisper-large-v3, whisper large v3

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works