openai
OpenAI: Whisper Large V3 Turbo
openai/whisper-large-v3-turbo
Our take
AI-assisted analysis based on the sources below. Suitability is an editorial assessment; test it on your own workload.
openai/whisper-large-v3-turbo is described by the model documentation as a fine-tuned, pruned Whisper large-v3 checkpoint for multilingual automatic speech recognition and translation into English. The documentation reports substantially faster inference than the unpruned checkpoint, with a minor quality trade-off, and support for automatic language detection, optional timestamps, batching, and long-audio processing in Transformers. OpenRouter's catalog separately lists audio-to-transcription availability. The supplied sources do not include benchmark results, so reported speed and quality effects should be validated on the intended workload.[1][2]
Strengths
- Supports multilingual speech recognition and speech translation into English, according to the model documentation.[1][2]
- The documentation reports substantially faster inference than Whisper large-v3 because decoding layers were pruned, while describing the quality loss as minor.[2]
- Transformers integration covers automatic language detection, transcription or English translation, optional timestamps, batching, and long-form audio.[2]
Limitations
- Pruning reduces decoder layers and the documentation says it causes minor quality degradation; no supplied source quantifies the accuracy impact for this exact variant.[1][2]
- The evidence pack contains a model card and an OpenRouter catalog entry, but no benchmark results or independent evaluation; reported behavior should be checked on target audio and languages.[1][2]
- Speech capabilities beyond transcription and English translation, such as speaker classification or diarization, are described as not robustly evaluated; contextual evaluation is advised before deployment.[2]
Where to use it
- Could be evaluated for high-volume transcription where faster inference is more important than avoiding the reported minor quality trade-off.[2]
- May suit multilingual transcription or speech-to-English translation features that benefit from automatic language detection and optional timestamps.[2]
- Could be considered for long recordings through the documented sequential or chunked long-form options, subject to validation of stitching and accuracy in the target domain.[2]
Sources
- [1] OpenRouter catalog · OpenAI: Whisper Large V3 TurboSources checked: Nov 13, 2026
- [2] Model documentation · OpenAI: Whisper Large V3 TurboSources checked: Nov 13, 2026
About this model
Description below is the original text supplied by OpenRouter, shown in the source language.
Whisper Large V3 Turbo is an optimized version of OpenAI's Whisper Large V3 speech recognition model, designed for speed and cost efficiency. It supports transcription across 99+ languages with a...
Specifications
- Model ID
- openai/whisper-large-v3-turbo
- Creator
- openai
- Added to OpenRouter
- 01.05.2026
- Context length
- 0
- Modality
- audio->transcription
- Input modalities
- audio
- Output modalities
- transcription
- Token pricing
- No per-token price in the source snapshot — see OpenRouter for the billing unit
- Knowledge cutoff
- —
- Expiration date
- —
Tasks and capabilities
A grade appears only where the source supplies a matching measurement for that task, the model is active and the cohort holds at least 10 comparable models. The capability column lists facts the source declares — supported formats, reasoning and tool parameters — which are not measured quality.
| Task | Grade | Source-declared capability |
|---|---|---|
| General quality | No measurement | — |
| Coding | No measurement | — |
| Agents | No measurement | — |
| Writing | No measurement | — |
| Reasoning | No measurement | — |
| Document analysis | No measurement | — |
| Vision | No measurement | — |
| Image generation | No measurement | — |
| Audio | No measurement | input: audio |
| Video | No measurement | — |
| Embeddings | No measurement | — |
Evaluations
No evaluations are supplied for this model.
How grades work
Each letter grade compares a model only against other models scored by the same source and metric in the current snapshot. S means the top 10%, A the next 20%, B the next 30%, C the rest. Models without a published measurement are shown as not rated — a supported format alone never implies measured quality.
- S — S — top 10% of the cohort
- A — A — next 20% of the cohort
- B — B — next 30% of the cohort
- C — C — below the B band
Providers
Optional live provider statistics from OpenRouter. Missing values are unknown, not zero.
| Provider | Context length | Price in /1M | Price out /1M | Latency p50 (30m, s) | Throughput p50 (30m, tok/s) | Uptime (30m) |
|---|---|---|---|---|---|---|
| DeepInfra | 0 | — | — | — | — | 100% |
| PrimeIntellectbf16 | 0 | — | — | — | — | 100% |
| Groq | 0 | — | — | — | — | 100% |
Provider data retrieved: 13.11.2026
Related articles
No published articles identify this model yet.