[Bug]: <title>Deepseek/deepseek-chat does not support response schemas (fails in create_community_reports workflow)

[zjfk474v6c-dev]

Do you need to file an issue?

• I have searched the existing issues and this bug is not already filed.
• My model is hosted on OpenAI or Azure. If not, please look at the "model providers" issue and don't file a new one here.
• I believe this is a legitimate bug, not just a question. If this is a question, please use the Discussions area.

Describe the bug

When running the graphrag indexing pipeline with deepseek/deepseek-chat as the default completion model, the create_community_reports workflow fails consistently. The core error is ValueError: Model 'deepseek/deepseek-chat' does not support response schemas, which prevents generating community reports for all clusters and eventually triggers a KeyError: 'community' due to missing report data during post-processing.
Environment
Graphrag version: Latest (from main branch)
Model: deepseek/deepseek-chat (via Litellm, API base: https://api.deepseek.com)
Pipeline configuration: Default setup with community_reports enabled
Python version: 3.12.12
OS: Windows 10

Steps to reproduce

• Configure graphrag to use deepseek/deepseek-chat as the default_completion_model (see config snippet below).
• Run the indexing pipeline with a sample input document.
• The pipeline proceeds normally through workflows like extract_graph and create_communities.
• Failure occurs when executing create_community_reports workflow.

Expected Behavior

No response

GraphRAG Config Used

settings.yaml

# Paste your config here

Logs and screenshots

indexing-engine.log

Additional Information

• GraphRAG Version:3.0.1
• Operating System:windows10
• Python Version:3.12
• Related Issues:

[IS430]

Same issue

[yteeh]

same with Qwen or ERNIE

[damianpodkonski]

same error with custom gpt 4.1-mini