langchain-ai/langgraph
1
0
Fork 0
mirror of https://github.com/langchain-ai/langgraph.git synced 2026-07-15 12:35:43 -04:00
Code Issues 314 Packages Projects Releases 17 Wiki Activity

usage_metadata is None when stream=True in ainvoke #913

New Issue
Closed
opened 2026-02-20 17:42:21 -05:00 by yindo · 3 comments
yindo commented 2026-02-20 17:42:21 -05:00
Owner
Copy Link
Copy Source

Originally created by @himanshu17287 on GitHub (Aug 12, 2025).

Checked other resources

  • This is a bug, not a usage question. For questions, please use the LangChain Forum (https://forum.langchain.com/).
  • I added a clear and detailed title that summarizes the issue.
  • I read what a minimal reproducible example is (https://stackoverflow.com/help/minimal-reproducible-example).
  • I included a self-contained, minimal example that demonstrates the issue INCLUDING all the relevant imports. The code run AS IS to reproduce the issue.

Example Code

response = await self.model.ainvoke(messages, config, stream=True)
print(response.usage_metadata)

Error Message and Stack Trace (if applicable)

Description

When using ainvoke with stream=True in LangGraph, the returned BaseMessage object’s usage_metadata field is always None.
In non-streaming mode (stream=False), usage_metadata is populated correctly.

System Info

langgraph = "^0.4.8"
langchain-openai = "^0.3.27"

Originally created by @himanshu17287 on GitHub (Aug 12, 2025). ### Checked other resources - [x] This is a bug, not a usage question. For questions, please use the LangChain Forum (https://forum.langchain.com/). - [x] I added a clear and detailed title that summarizes the issue. - [x] I read what a minimal reproducible example is (https://stackoverflow.com/help/minimal-reproducible-example). - [x] I included a self-contained, minimal example that demonstrates the issue INCLUDING all the relevant imports. The code run AS IS to reproduce the issue. ### Example Code ```python response = await self.model.ainvoke(messages, config, stream=True) print(response.usage_metadata) ``` ### Error Message and Stack Trace (if applicable) ```shell ``` ### Description When using ainvoke with stream=True in LangGraph, the returned BaseMessage object’s usage_metadata field is always None. In non-streaming mode (stream=False), usage_metadata is populated correctly. ### System Info langgraph = "^0.4.8" langchain-openai = "^0.3.27"
yindo added the bugpendingawaiting author revision labels 2026-02-20 17:42:21 -05:00
yindo closed this issue 2026-02-20 17:42:21 -05:00
yindo commented 2026-02-20 17:42:21 -05:00
Author
Owner
Copy Link
Copy Source

@eyurtsev commented on GitHub (Aug 12, 2025):

Hi @himanshu17287 could you include a full MRE? which model are you using what does your code look like?

@eyurtsev commented on GitHub (Aug 12, 2025): Hi @himanshu17287 could you include a full MRE? which model are you using what does your code look like?
yindo commented 2026-02-20 17:42:21 -05:00
Author
Owner
Copy Link
Copy Source

@himanshu17287 commented on GitHub (Aug 18, 2025):

Hi @eyurtsev

Model name : gpt-4o

self.model = await ModelsProvider().get_chat_model("gpt-4o")

response = await self.model.ainvoke(
messages, config, stream=true
)
print(response.usage_metadata)

When I set stream = false, I am able to fetch usage metadata, but not when it is set to true

@himanshu17287 commented on GitHub (Aug 18, 2025): Hi @eyurtsev Model name : gpt-4o self.model = await ModelsProvider().get_chat_model("gpt-4o") response = await self.model.ainvoke( messages, config, stream=true ) print(response.usage_metadata) When I set stream = false, I am able to fetch usage metadata, but not when it is set to **true**
yindo commented 2026-02-20 17:42:21 -05:00
Author
Owner
Copy Link
Copy Source

@casparb commented on GitHub (Sep 15, 2025):

Hi @himanshu17287. OpenAI's Chat Completions API does not stream token usage statistics by default. To recover token counts when streaming with OpenAI models, set stream_usage=True as an initialization parameter or on invocation:

from langchain_openai import ChatOpenAI

llm = ChatOpenAI(model="gpt-4o", stream_usage=True)

response = await llm.ainvoke("What is the capital of France?", stream=True)

See the docs.

@casparb commented on GitHub (Sep 15, 2025): Hi @himanshu17287. OpenAI's Chat Completions API does not stream token usage statistics by default. To recover token counts when streaming with OpenAI models, set `stream_usage=True` as an initialization parameter or on invocation: ```python from langchain_openai import ChatOpenAI llm = ChatOpenAI(model="gpt-4o", stream_usage=True) response = await llm.ainvoke("What is the capital of France?", stream=True) ``` [See the docs](https://docs.langchain.com/oss/python/integrations/chat/openai#streaming-usage-metadata).
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
agents
awaiting author revision
bug
change
chat models
checkpointer
ci
code quality
code quality
dependencies
documentation
drawing
duplicate
enhancement
good first issue
help wanted
invalid
maintainer
meta
missing information
not maintained by langchain
open-swe-auto
open-swe-max
open-swe-max-auto
pending
performance
pull-request
Mirrored from GitHub Pull Request
 question
requires design
sdk
server
stale
studio
tests
type checking
usage question
v0.5
v1.1
No Label awaiting author revision bug pending
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: langchain-ai/langgraph#913
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?