docling-project/docling · error · ValueError

RapidOCR backend does not support language . Supported: .

Error message

RapidOCR {backend} backend does not support language {lang!r}. Supported: {sorted(PP_OCRV6_LANGS | _PPOCRV5_LANGS)}.

What it means

For non-torch backends the language resolver falls back from PP-OCRv6 to PP-OCRv5. If the requested language (after normalization and aliasing) is in neither PP_OCRV6_LANGS nor _PPOCRV5_LANGS, this ValueError is raised listing the supported codes for this backend.

Solutions

1. Use one of the language codes listed in the error message (they are the exact accepted tokens after aliasing).
2. Check Docling's _DOCLING_LANG_NORMALIZE and rapidocr's COMMON_LANG_ALIASES to confirm your spelling maps correctly.
3. If the language is unsupported by RapidOCR entirely, select a different OCR engine via pipeline_options.ocr_options.

Example fix

# before
ocr_options = RapidOcrOptions(lang="jpan")

# after
ocr_options = RapidOcrOptions(lang="japan")

from rapidocr.utils.model_resolver import COMMON_LANG_ALIASES, PP_OCRV6_LANGS

code = lang.strip().lower()
aliased = COMMON_LANG_ALIASES.get(code, code)
assert aliased in PP_OCRV6_LANGS | _PPOCRV5_LANGS, f"unsupported lang {lang!r}"

def is_supported_rapidocr_lang(lang: str) -> bool:
    from rapidocr.utils.model_resolver import COMMON_LANG_ALIASES, PP_OCRV6_LANGS
    code = lang.strip().lower()
    aliased = COMMON_LANG_ALIASES.get(code, code)
    return aliased in PP_OCRV6_LANGS  # v6 is the preferred superset

try:
    RapidOcrModel(options=RapidOcrOptions(lang=lang, backend=backend))
except ValueError as e:
    if "does not support language" in str(e):
        log.warning("unsupported lang %r, falling back to default OCR language", lang)
        options.lang = ["en"]
    else:
        raise

When it happens

Trigger: RapidOcrOptions(lang=X, backend='onnxruntime'|'openvino'|'paddle') where X is not a PP-OCRv5 or PP-OCRv6 language code, e.g. a fabricated code like 'klingon' or an unsupported regional code.

Common situations: Misspelled language codes ('ch' vs 'chinese'); using ISO codes that rapidocr does not alias; assuming a language is supported because another OCR engine supports it.

Related errors

• Invalid RapidOCR model spec
• Invalid RapidOCR model spec
• RapidOCR torch backend does not support language
• Unknown RapidOCR backend
• Unknown RapidOCR backend

AI-assisted analysis of docling-project/docling@61d76f1ff3 (2026-08-14). Data as JSON: /api/errors/470ea162cf7d9b9e. Report an issue: GitHub.

Appendix: source

Thrown at docling/models/stages/ocr/rapid_ocr_model.py:168

    from rapidocr.utils.typings import OCRVersion

    code = lang.strip().lower()
    code = _DOCLING_LANG_NORMALIZE.get(code, code)
    aliased = COMMON_LANG_ALIASES.get(code, code)

    if aliased in PP_OCRV6_LANGS:
        version = OCRVersion.PPOCRV6
    elif backend == "torch":
        if aliased not in _PPOCRV4_LANGS:
            raise ValueError(
                f"RapidOCR torch backend does not support language {lang!r}. "
                f"Supported: {sorted(PP_OCRV6_LANGS | _PPOCRV4_LANGS)}."
            )
        version = OCRVersion.PPOCRV4
    elif aliased in _PPOCRV5_LANGS:
        version = OCRVersion.PPOCRV5
    else:
        raise ValueError(
            f"RapidOCR {backend} backend does not support language {lang!r}. "
            f"Supported: {sorted(PP_OCRV6_LANGS | _PPOCRV5_LANGS)}."
        )

    _log.debug(
        "RapidOCR resolved lang=%r backend=%r -> version=%s rec_lang=%r",
        lang,
        backend,
        version.value,
        aliased,
    )
    return _RapidOcrModelSpec(
        backend=backend,
        user_lang=lang,
        rapidocr_lang_token=aliased,
        ppocr_version=version,
    )

View on GitHub (pinned to 61d76f1ff3)