Skip to content

docs(sdk): clarify Responses streaming and tokenizer boundaries - #825

Closed
vedjoshi1 wants to merge 1 commit into
OpenHands:mainfrom
vedjoshi1:refactor/4976-typed-llm-boundaries
Closed

vedjoshi1 wants to merge 1 commit into
OpenHands:mainfrom
vedjoshi1:refactor/4976-typed-llm-boundaries

Conversation

@vedjoshi1

@vedjoshi1 vedjoshi1 commented Sep 23, 2026

Copy link
Copy Markdown
  • I have read and reviewed the documentation changes to the best of my ability.
  • If the change is significant, I have run the documentation site locally and confirmed it renders as expected.

Summary of changes

This accompanies SDK issue #4976
and the SDK implementation PR.

  • Document Responses API streaming through responses() and aresponses(),
    callback behavior, and final completion selection.
  • Describe the private provider boundaries and optional chat-template tokenizer
    fallback in the SDK architecture page.
  • Remove the outdated statement that only Chat Completions supports streaming.

Verification

uv run --with pytest --with requests --with pyyaml pytest -q tests/ produced
35 passed and 2 unrelated setup errors. Both errors are from the existing
pricing tests fetching OpenHands/OpenHands/main/openhands/utils/llm.py, which
currently returns HTTP 404. No pricing files or test fixtures changed.

The paired SDK branch passed 1,073 LLM/checker tests, 6,432 SDK tests, and six
local real HTTP/SSE cases. It also built and started the packaged Agent Server.

Co-authored-by: openhands <openhands@all-hands.dev>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant