langchain-ai/langchain · error · ValueError

Unsupported output parser {output_parser_type}

Error message

Unsupported output parser {output_parser_type}

What it means

`_load_output_parser` only supports output parsers whose config `_type` equals `"default"` (mapped to `StrOutputParser`). Any other `_type` in the `output_parser` section of a prompt config raises `ValueError: Unsupported output parser {type}`.

Source

Thrown at libs/core/langchain_core/prompts/loading.py:147

                examples = json.load(f)
            elif path.suffix in {".yaml", ".yml"}:
                examples = yaml.safe_load(f)
            else:
                msg = "Invalid file format. Only json or yaml formats are supported."
                raise ValueError(msg)
        config["examples"] = examples
    else:
        msg = "Invalid examples format. Only list or string are supported."
        raise ValueError(msg)  # noqa:TRY004
    return config


def _load_output_parser(config: dict[str, Any]) -> dict[str, Any]:
    """Load output parser."""
    if config_ := config.get("output_parser"):
        if output_parser_type := config_.get("_type") != "default":
            msg = f"Unsupported output parser {output_parser_type}"
            raise ValueError(msg)
        config["output_parser"] = StrOutputParser(**config_)
    return config


def _load_few_shot_prompt(
    config: dict[str, Any], *, allow_dangerous_paths: bool = False
) -> FewShotPromptTemplate:
    """Load the "few shot" prompt from the config."""
    # Load the suffix and prefix templates.
    config = _load_template(
        "suffix", config, allow_dangerous_paths=allow_dangerous_paths
    )
    config = _load_template(
        "prefix", config, allow_dangerous_paths=allow_dangerous_paths
    )
    # Load the example prompt.
    if "example_prompt_path" in config:
        if "example_prompt" in config:

View on GitHub (pinned to e32fa9a52e)

Solutions

1. Remove the `output_parser` section from the config and attach the parser in code after loading: `prompt | MyParser()`.
2. Change `_type` to `"default"` if a plain string parser is acceptable.
3. Migrate to `langchain_core.load.dumpd`/`dumps` + `load`/`loads` for full round-trip serialization.

Example fix

# before
# {"_type": "prompt", "template": "...",
#  "output_parser": {"_type": "regex", "pattern": "..."}}
load_prompt('prompt.json')  # ValueError

# after
prompt = load_prompt('prompt.json')  # parser removed from config
chain = prompt | RegexParser(pattern="...")

op = config.get('output_parser')
if op and op.get('_type', 'default') != 'default':
    del config['output_parser']  # reattach in code instead
load_prompt_from_config(config)

def is_default_output_parser(config: dict) -> bool:
    op = config.get('output_parser')
    return op is None or op.get('_type') == 'default'

When it happens

Trigger: A prompt config containing `"output_parser": {"_type": "regex", ...}` or `{"_type": "structured"}`, ... loaded via `load_prompt`.

Common situations: Legacy langchain configs that serialized `RegexParser`, `PydanticOutputParser`, etc.; Hub-era prompt files with parser sections; assuming core's loader covers all legacy parser types.

Related errors

• Loading {config_type} prompt not supported
• Both `{var_name}_path` and `{var_name}` cannot be provided.
• Unsupported template file format: '{resolved_path.suffix}'.
• Invalid examples format. Only list or string are supported.
• Only one of example_prompt and example_prompt_path should be

AI-assisted analysis of langchain-ai/langchain@e32fa9a52e (2026-08-14). Data as JSON: /api/errors/c7b00ac7f9bec05f. Report an issue: GitHub.