[open-swe] feat: Improve token usage tracking documentation with LLM and chat model approaches

[open-swe]

Fixes #8491

Overview

This PR enhances the token usage tracking documentation by:

• Clarifying the difference between LLM and chat model token usage tracking
• Providing clear guidance on accessing token usage for different model types
• Adding new example files to demonstrate direct and callback-based approaches

Key Changes

• Updated docs/core_docs/docs/how_to/llm_token_usage_tracking.mdx:

  • Added comprehensive explanation of LLM vs Chat model token usage
  • Restructured content to provide clear guidance
  • Added info box explaining token usage differences

• Created new example file examples/src/models/llm/token_usage_tracking_direct.ts:

  • Demonstrates direct token usage tracking for LLMs
  • Shows how to access llmOutput.tokenUsage

• Updated examples/src/models/llm/token_usage_tracking.ts:

  • Refocused as alternative callback-based approach
  • Added explanatory comments

Rationale

The documentation now provides clear, comprehensive guidance on token usage tracking for both LLMs and chat models, addressing potential confusion about different tracking methods.

Testing

• Verified documentation accuracy
• Confirmed example files run without errors
• Ensured clear explanations of token usage approaches

Resolves the original request to improve token usage tracking documentation.

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Comments	Updated (UTC)
langchainjs-docs	🛑 Canceled (Inspect)		💬 Add feedback	Jul 17, 2025 5:44pm

1 Skipped Deployment

Name	Status	Preview	Comments	Updated (UTC)
langchainjs-api-refs	⬜️ Ignored (Inspect)			Jul 17, 2025 5:44pm