PrefectHQ/fastmcp · error · ValueError

Unsupported audio MIME type for OpenAI: {content.mime_type!r

Error message

Unsupported audio MIME type for OpenAI: {content.mime_type!r}. Supported types: {', '.join(sorted(_OPENAI_AUDIO_FORMATS))}

What it means

OpenAI audio input accepts only wav and mp3 (see _OPENAI_AUDIO_FORMATS). When converting MCP AudioContent with any other MIME type, the handler raises ValueError enumerating the supported formats.

Source

Thrown at fastmcp_slim/fastmcp/client/sampling/handlers/openai.py:86

    if content.mime_type not in _OPENAI_IMAGE_MEDIA_TYPES:
        raise ValueError(
            f"Unsupported image MIME type for OpenAI: {content.mime_type!r}. "
            f"Supported types: {', '.join(sorted(_OPENAI_IMAGE_MEDIA_TYPES))}"
        )
    data_url = f"data:{content.mime_type};base64,{content.data}"
    return ChatCompletionContentPartImageParam(
        type="image_url",
        image_url={"url": data_url},
    )


def _audio_content_to_openai_part(
    content: AudioContent,
) -> ChatCompletionContentPartInputAudioParam:
    """Convert MCP AudioContent to OpenAI input_audio content part."""
    audio_format = _OPENAI_AUDIO_FORMATS.get(content.mime_type)
    if audio_format is None:
        raise ValueError(
            f"Unsupported audio MIME type for OpenAI: {content.mime_type!r}. "
            f"Supported types: {', '.join(sorted(_OPENAI_AUDIO_FORMATS))}"
        )
    return ChatCompletionContentPartInputAudioParam(
        type="input_audio",
        input_audio={"data": content.data, "format": audio_format},
    )


class OpenAISamplingHandler:
    """Sampling handler that uses the OpenAI API."""

    def __init__(
        self,
        default_model: ChatModel,
        client: AsyncOpenAI | None = None,
    ) -> None:
        self.client: AsyncOpenAI = client or AsyncOpenAI()

View on GitHub (pinned to 1f02114297)

Solutions

Transcode the audio to WAV or MP3 and update mime_type to 'audio/wav' or 'audio/mp3'.
Route sampling to a handler supporting the format (e.g. google-genai).
Set the mime_type explicitly if it was omitted.
Skip/replace the AudioContent if audio isn't essential to the request.

Example fix

// before
AudioContent(type='audio', data=ogg_b64, mime_type='audio/ogg')
// after
AudioContent(type='audio', data=wav_b64, mime_type='audio/wav')

Defensive patterns

Strategy: validation

Validate before calling

for m in messages:
    for c in getattr(m, 'content', []):
        if getattr(c, 'type', None) == 'audio' and c.mime_type not in ('audio/wav', 'audio/mp3'):
            raise ValueError(f"Transcode {c.mime_type} to wav/mp3 before OpenAI sampling")

Type guard

def is_openai_supported_audio(c) -> bool:
    return getattr(c, 'type', None) == 'audio' and c.mime_type in ('audio/wav', 'audio/mp3')

Try / catch

try:
    result = await client.sample(...)
except ValueError as e:
    if 'Unsupported audio MIME type for OpenAI' in str(e):
        messages = transcode_audio_to_wav(messages)
        result = await client.sample(messages=messages, ...)
    else:
        raise

Prevention

Transcode browser recordings (ogg/webm) to wav/mp3 before sampling.
Always set mime_type explicitly on AudioContent.
Route unsupported audio formats to a different provider handler.

When it happens

Trigger: Client.sample() with AudioContent whose mime_type maps to nothing in _OPENAI_AUDIO_FORMATS (e.g. 'audio/ogg', 'audio/webm', 'audio/flac', or missing mime_type), routed through _convert_to_openai_messages -> _audio_content_to_openai_part.

Common situations: Recordings captured as ogg/webm (common from browsers) passed to OpenAI sampling; transcription pipelines using flac; AudioContent constructed without mime_type.

Related errors

AI-assisted analysis of PrefectHQ/fastmcp@1f02114297 (2026-08-29). Data as JSON: /api/errors/967e242a61105d59. Report an issue: GitHub.