ScrapeGraphAI/Scrapegraph-ai · error · ValueError

Could not determine model name from llm_model. Please specif

Error message

Could not determine model name from llm_model. Please specify 'model' in batch_config.

What it means

BatchGenerateAnswerNode._get_model_name tries to read the model name from the LangChain model instance via 'model_name' then 'model' attributes; if neither exists it raises this error telling you to set 'model' explicitly in batch_config. This name is needed to build batch API requests.

Source

Thrown at scrapegraphai/nodes/batch_generate_answer_node.py:95

        self.batch_model = batch_config.get("model")
        self.batch_temperature = batch_config.get("temperature", 0.0)

    def _get_model_name(self) -> str:
        """Extract the OpenAI model name from the LLM configuration.

        Returns:
            The model name string (e.g., 'gpt-4o-mini').
        """
        if self.batch_model:
            return self.batch_model

        # Try to extract model name from the LangChain model instance
        if hasattr(self.llm_model, "model_name"):
            return self.llm_model.model_name
        if hasattr(self.llm_model, "model"):
            return self.llm_model.model

        raise ValueError(
            "Could not determine model name from llm_model. "
            "Please specify 'model' in batch_config."
        )

    def _get_format_instructions(self) -> str:
        """Get format instructions based on the schema configuration."""
        if self.schema is not None:
            output_parser = get_pydantic_output_parser(self.schema)
            return output_parser.get_format_instructions()
        return (
            "You must respond with a JSON object. Your response should be "
            "formatted as a valid JSON with a 'content' field containing "
            'your analysis. For example:\n'
            '{"content": "your analysis here"}'
        )

    def _build_prompt_text(
        self,

View on GitHub (pinned to 532dfffbf6)

Solutions

1. Set the model name explicitly in node config: node_config={'batch_config': {'model': 'gpt-4o-mini'}}
2. Check dir(llm_model) to see which attribute holds the model name and use a model class that exposes model_name or model
3. Subclass BatchGenerateAnswerNode and override _get_model_name for custom model wrappers

Example fix

# before
node = BatchGenerateAnswerNode(node_config={'batch_config': {}})
# after
node = BatchGenerateAnswerNode(
    node_config={'batch_config': {'model': 'gpt-4o-mini'}}
)

def model_name_available(llm) -> bool:
    return hasattr(llm, 'model_name') or hasattr(llm, 'model')

# or just set it explicitly:
node_config = {'batch_config': {'model': 'gpt-4o-mini'}}

from typing import Protocol
class NamedModel(Protocol):
    model_name: str

def has_model_name(m) -> bool:
    return hasattr(m, 'model_name') or hasattr(m, 'model')

try:
    name = node._get_model_name()
except ValueError:
    node.node_config.setdefault('batch_config', {})['model'] = 'gpt-4o-mini'

When it happens

Trigger: Passing an llm_model wrapper that exposes neither .model_name nor .model (some third-party/custom LangChain integrations); constructing the node with a client object instead of a chat model instance.

Common situations: Using a newer LangChain integration that renamed the attribute; passing Azure/other providers whose attribute naming differs; forgetting to set batch_config={'model': ...} when using an exotic model.

Related errors

• No parsed documents found in state
• Response took longer than {timeout} seconds
• Model not supported
• model_tokens not specified
• Provider {llm_params["model_provider"]} is not supported.

AI-assisted analysis of ScrapeGraphAI/Scrapegraph-ai@532dfffbf6 (2026-08-28). Data as JSON: /api/errors/9f2e041cf44ffb0b. Report an issue: GitHub.