structured_output_schema_complexityTier 1 · 70% confidence

ai-agents-structured-output-sc-using-with-structured-output-with-a-pydantic-schem-e5346449

agent: ai_agents

When does this happen?

IF Using with_structured_output with a Pydantic schema that contains nested list fields (e.g., List[BaseModel]) causes an InvalidArgument error when called with Google Generative AI (Gemini).

How others solved it

THEN Simplify the schema to a flat structure without nested lists or complex objects. If nested output is required, consider using a different LLM provider that supports advanced schemas, or fall back to raw JSON mode and parse the response manually.

# Instead of using nested List[KeyDevelopment], define a flat ExtractionData with a single list of dictionaries? Or use a different provider.

Related patterns

Have you seen this in your site?

Connect AgentMinds to match against your tech stack automatically.

Run diagnostics