Aider-AI/aider · critical · ValueError

Please set the OPENAI_API_KEY environment variable.

Error message

Please set the OPENAI_API_KEY environment variable.

What it means

The __main__ block of aider/voice.py requires OPENAI_API_KEY before constructing Voice(), because transcription uses OpenAI's audio API. It raises ValueError('Please set the OPENAI_API_KEY environment variable.') when os.getenv returns empty. Note this is script-invocation-only (python -m aider.voice or python aider/voice.py); when aider runs normally, voice failures surface as different errors from the configured model.

Source

Thrown at aider/voice.py:186

            try:
                transcript = litellm.transcription(
                    model="whisper-1", file=fh, prompt=history, language=language
                )
            except Exception as err:
                print(f"Unable to transcribe {filename}: {err}")
                return

        if filename != temp_wav:
            os.remove(filename)

        text = transcript.text
        return text


if __name__ == "__main__":
    api_key = os.getenv("OPENAI_API_KEY")
    if not api_key:
        raise ValueError("Please set the OPENAI_API_KEY environment variable.")
    print(Voice().record_and_transcribe())

View on GitHub (pinned to 5dc9490bb3)

Solutions

1. Export the key: export OPENAI_API_KEY=sk-... (or add it to your shell profile) and re-run.
2. For persistent use, store it in your shell rc or a secrets manager that injects env vars.
3. If you use a non-OpenAI provider, invoke voice through the main aider CLI with your configured model rather than this script entry point.

Example fix

# before
$ python aider/voice.py  # ValueError: Please set the OPENAI_API_KEY environment variable.

# after
$ export OPENAI_API_KEY="sk-..."
$ python aider/voice.py

import os, sys

if not os.getenv("OPENAI_API_KEY"):
    sys.exit("OPENAI_API_KEY not set — export it before using voice transcription.")

When it happens

Trigger: Running aider/voice.py directly as a script without OPENAI_API_KEY exported. Checked at startup via os.getenv, so it fails before any recording or transcription attempt.

Common situations: Quick mic-transcription tests with python aider/voice.py in a fresh shell where the key lives in another profile's rc file; CI/scheduled environments without the secret; users on Azure/local models who never set an OpenAI key.

Related errors

• summarizer unexpectedly failed for all models
• GITHUB_COPILOT_TOKEN environment variable not found
• GITHUB_COPILOT_TOKEN environment variable is empty
• Device '{device_name}' not found. Available input devices: {
• Unsupported audio format: {audio_format}

AI-assisted analysis of Aider-AI/aider@5dc9490bb3 (2026-08-15). Data as JSON: /api/errors/21a1d6a575844664. Report an issue: GitHub.