fishaudio/fish-speech · error · HTTPException

No audio generated, please check the input text.

Error message

No audio generated, please check the input text.

What it means

After consuming the whole result stream, if zero audio segments were yielded the wrapper raises HTTP 500 'No audio generated...'. This means inference ran but produced no audible output, typically because the text produced no valid tokens.

Source

Thrown at tools/server/inference.py:42

            case "error":
                raise HTTPException(
                    HTTPStatus.INTERNAL_SERVER_ERROR,
                    content=str(result.error),
                )

            case "segment":
                count += 1
                if isinstance(result.audio, tuple):
                    yield (result.audio[1] * AMPLITUDE).astype(np.int16).tobytes()

            case "final":
                count += 1
                if isinstance(result.audio, tuple):
                    yield result.audio[1]
                return None  # Stop the generator

    if count == 0:
        raise HTTPException(
            HTTPStatus.INTERNAL_SERVER_ERROR,
            content="No audio generated, please check the input text.",
        )

View on GitHub (pinned to befe400174)

Solutions

1. Check the request text is non-empty and contains real words
2. Use a model/checkpoint matching the input language
3. Try the same text via the CLI to see model-side logs; slightly lengthen the input

text = (req.text or "").strip()
if not text or not any(c.isalpha() for c in text):
    raise ValueError("text must contain actual words")

if resp.status_code == 500 and "No audio generated" in resp.text:
    text = add_filler(text)  # retry with longer, richer input

When it happens

Trigger: Sending empty/whitespace text, text in a language the model can't handle, or content that tokenizes to nothing usable — the stream completes without any 'segment' result.

Common situations: Passing punctuation-only strings; wrong language for the checkpoint; very short inputs the codec drops; silent failures after fine-tuning.

Related errors

• {str(result.error)}
• Failed to encode audio
• Failed to decode tokens to audio
• Invalid token
• Unsupported Media Type

AI-assisted analysis of fishaudio/fish-speech@befe400174 (2026-08-27). Data as JSON: /api/errors/28830ae77b6227de. Report an issue: GitHub.