NEWVectors or files. Pick a path.Start →
    Provider APITranscriptionVerified 2026-11-05

    OpenAI Whisper API (whisper-1)

    whisper-1

    OpenAI Whisper API (whisper-1) is a audio and video speech-to-text model (OpenAI API terms of use). Called on Mixpeek through the provider API.

    Mixpeek calls the provider's API by name. Provider terms apply.

    Use in a new namespaceAdd to an existing namespace Record as JSON

    New opens the Designer in Studio. For an existing namespace, open it in Studio and use its Designer tab.

    No demo for OpenAI Whisper API (whisper-1) yet

    A video, image, slideshow or playground goes here once we record one on our own data.

    What it is good and bad at

    Best at
    Speech-to-text for video and audio with no GPU in the cluster; the production default for multimodal_extractor transcription.
    Worst at
    Provider rate limits (a Primis batch lost its transcripts to a 429), per-minute cost, and audio sent off-box.
    Task
    Transcription
    Input
    Audio, Video
    Output
    Text
    Use cases
    Transcripts

    Templates that use it

    No template uses this model yet. Browse templates.

    Input and output

    Input

    audio
    audio chunks from AudioParallelChunking
    language
    optional forced language

    Output

    text
    transcript, or verbose_json segments when segment timestamps are wanted

    How Mixpeek wires it

    Extractors
    • multimodal_extractor@v1 run_transcription=true and use_whisper_api (HARDWARE_WHISPER_USE_API=true, or forced when the cluster has no GPU)
    • multimodal_extractor@v2 run_transcription=true and use_whisper_api (HARDWARE_WHISPER_USE_API=true, or forced when the cluster has no GPU)
    Feature URIs
    None today
    Vector indexes
    None
    Retriever stages
    None
    Combines with
    • video -> audio chunks -> whisper-1 transcript -> E5 transcription vector

    How it runs

    Status
    Provider API OpenAI audio transcriptions API, called from WhisperTranscriptionBatch; universal__api also maps whisper-1 for POST /v1/inference.
    Serve apps
    universal__api
    Runtime
    OpenAI audio.transcriptions.create (engine.inference.openai.whisper_api.services)
    Weights
    Provider API, no remote code
    Parameters
    Not recorded
    Size on disk
    Not recorded
    Preprocessing
    ffmpeg audio extraction and chunking per segment.
    Postprocessing
    Text, or verbose_json segments mapped to timestamps.
    Serving
    both
    Hardware
    Provider API

    Scale to zero

    Minimum replicas
    0
    Maximum replicas
    6
    Idle timeout
    600 s
    Memory
    Not recorded
    CPUs
    0.1
    GPUs
    0
    Cold start, node warm
    Not recorded
    Cold start, new node
    Not recorded

    Measured cost

    Unmeasured

    Billed per audio minute by OpenAI; no committed measurement.

    Our evaluations

    Unmeasured

    We have not evaluated this model on our own data yet. Publisher results are in the sources below.

    Questions

    What is OpenAI Whisper API (whisper-1) best at?
    Speech-to-text for video and audio with no GPU in the cluster; the production default for multimodal_extractor transcription.
    What are the limits of OpenAI Whisper API (whisper-1)?
    Provider rate limits (a Primis batch lost its transcripts to a 429), per-minute cost, and audio sent off-box.
    Can OpenAI Whisper API (whisper-1) be used commercially?
    Yes. Paid OpenAI API under OpenAI's terms.
    How does OpenAI Whisper API (whisper-1) run on Mixpeek?
    OpenAI audio transcriptions API, called from WhisperTranscriptionBatch; universal__api also maps whisper-1 for POST /v1/inference.