deepagents==0.4.0

langchain-ai/deepagentsdeepagents==0.4.0Feb 10, 2026by github-actions[bot]

AI Summary

Introduces new sandbox integration packages and refactors the SDK's summarization middleware to handle context overflows and utilize the Responses API.

Key Highlights

New sandbox packages: langchain-modal, langchain-daytona, and langchain-runloop.
Refactored summarization middleware now retains full message history in the graph state.
Exception handling for context overflows in ChatOpenAI and ChatAnthropic models.
Default to the Responses API when model strings are prefixed with 'openai:'.

New Features

New sandbox integration packages
Summarization middleware refactor
Context overflow exception handling
Responses API defaulting

Full Release Notes

Changes since deepagents==0.3.12

New integration packages for pluggable sandboxes: [langchain-modal](https://pypi.org/project/langchain-modal/), [langchain-daytona](https://pypi.org/project/langchain-daytona/), and [langchain-runloop](https://pypi.org/project/langchain-daytona/). See [sandboxes guide](/oss/deepagents/sandboxes) and example [data analysis tutorial](/oss/deepagents/data-analysis).

Notable changes from #1234
* Refactor the SDK's summarization middleware.
  * Summarization now happens in the model node via wrap_model_call events. Due to this we retain the full message history in the graph state.
  * System messages and tool definitions are accurately counted toward summarization triggers.
  * Add exception handling for context overflows: `ChatOpenAI` and `ChatAnthropic` now raise a standard LangChain exception when input exceeds a model's token limit. We catch this error and trigger summarization.
* When model strings in `create_deep_agent` are prefixed with `'openai:'`, we default to the Responses API.

chore(sdk): add support for `,` in allowed-tools, fail gently for bullet list (#1235)
fix(sdk): relax skills parsing for allowed tools (#1232)
chore(sdk)!: move sandbox provider back to cli (#1226)
refactor(sdk): move function-level test imports out to top (#1168)
fix(sdk): harden skills metadata parsing, improve spec compliance (#1189)
style: update brand names (#1180)

---

Thanks to our contributors: @sydney-runkle, @eyurtsev, @mdrxy