GenAI PMdeepagents
An OS-based agent framework referenced as portable across runtimes. The newsletter emphasizes that it can run in multiple environments without runtime lock-in.
Key Highlights
- DeepAgents is positioned as an OS-based, portable agent harness rather than a runtime-locked platform.
- It evolved from an open-source Claude Agent SDK into a broader reference point for production agent workflows.
- Key themes include dynamic subagents, middleware extensibility, large-scale evaluation, and OSS model optimization.
- For AI PMs, DeepAgents is especially relevant for reducing lock-in, designing long-horizon agents, and planning evaluation strategy.
- Newsletter mentions connect DeepAgents to LangSmith, LangGraph, Temporal, Gemini Live, and other major agent ecosystem tools.
Overview
DeepAgents is an open-source, OS-based agent framework introduced by Harrison Chase and associated with the Claude Agent SDK. Across newsletter mentions, it is positioned less as a single locked platform and more as a portable agent harness that can run across different runtimes and environments. That portability is a central theme: DeepAgents is described as something you can use in SuperQode, LangGraph, Temporal, or other platforms without being tied to one runtime stack.
For AI Product Managers, DeepAgents matters because it reflects a broader industry shift from lightweight orchestration frameworks toward more complete agent harnesses. It sits at the intersection of agent runtime design, evaluation, subagent orchestration, voice interfaces, middleware extensibility, and cost optimization for OSS models. In practice, that makes it relevant for PMs thinking about long-horizon workflows, platform portability, vendor risk, and how to operationalize agents beyond demos.
Key Developments
- 2026-01-14: DeepAgents was referenced as a capability inside LangSmith Agent Builder GA, alongside memory, MCPS/skills/subagents, triggers, and an agent inbox for human review.
- 2026-04-10: Harrison Chase noted that community middleware such as langchain-task-steering was emerging to customize agents and DeepAgents, signaling an extensibility ecosystem.
- 2026-04-13: DeepAgents was launched as an open-source Claude Agent SDK at `github.com/langchain-ai/deepagents`, positioned for building and running Claude-powered agents.
- 2026-04-30: Harrison Chase said he was optimizing DeepAgents for strong performance on OSS models, reflecting a cost and model-portability strategy.
- 2026-06-01: He outlined how to evaluate DeepAgents at scale on AWS with LangSmith, including datapoint and evaluator design for longer-horizon agents.
- 2026-06-21: DeepAgents appeared as part of a broad agentic AI learning stack alongside LangChain, LangGraph, RAG, and guardrails, indicating growing educational relevance.
- 2026-06-30: Dynamic subagents were introduced in DeepAgents, enabling developers to programmatically spin up subagents for multiple use cases.
- 2026-07-01: Harrison Chase showed a live voice agent pattern that used Gemini Live for low-latency interaction while offloading deeper reasoning to DeepAgents.
- 2026-07-06: He framed the market as moving from frameworks like LangChain, AI SDK, and LlamaIndex toward fuller harnesses such as DeepAgents, Claude Agent SDK, and EVE.
- 2026-07-10: Chase explicitly clarified that DeepAgents is not runtime lock-in; because it is OS-based, it can run across environments including SuperQode, LangGraph, Temporal, and others.
Relevance to AI PMs
1. Plan for portability and reduce platform risk
DeepAgents is repeatedly framed as portable across runtimes. PMs evaluating agent infrastructure can use this as a lens for avoiding vendor lock-in, preserving migration options, and designing agent products that can move between orchestration environments as requirements change.
2. Design for production-grade agent workflows, not just prompts
The mentions around dynamic subagents, memory, triggers, middleware, and evaluation suggest DeepAgents is meant for multi-step, longer-horizon workflows. PMs can use these capabilities to scope more realistic agent products with decomposition, human review, and extensibility built in.
3. Operationalize evaluation, cost, and model strategy early
DeepAgents is linked to evaluation at scale with LangSmith and AWS, and to optimization for OSS models. For PMs, this highlights a practical playbook: define evaluator frameworks early, benchmark long-running tasks, and keep model strategy flexible as costs and performance shift.
Related
- Harrison Chase: Primary advocate and the main source of public framing around DeepAgents.
- Claude Agent SDK / claude-agent-sdk: DeepAgents was launched as an open-source Claude Agent SDK and is closely identified with that label.
- Anthropic / Claude: DeepAgents initially centered on Claude-powered agents.
- LangSmith / langsmith-agent-builder: Connected through no-code agent building, evaluation, and operational tooling.
- LangChain / LangGraph: Frequently mentioned alongside DeepAgents; LangGraph is also cited as one environment where DeepAgents can run.
- Temporal: Mentioned as another platform/runtime context compatible with DeepAgents.
- SuperQode: Used as an example to show DeepAgents can run with a different runtime and is not platform-locked.
- memory: Appears as a complementary capability in agent systems that include DeepAgents.
- mcpsskillssubagents: Related to the modular agent tooling stack referenced with DeepAgents in LangSmith Agent Builder.
- guardrails: Mentioned in educational context as part of the broader production agent stack.
- Gemini Live: Paired with DeepAgents in a voice agent architecture where live interaction and deeper reasoning are split.
- AI SDK / LlamaIndex / EVE: Comparative reference points in the market shift from frameworks to more complete agent harnesses.
- OSS models: Important to DeepAgents’ cost and performance positioning.
- langchain-task-steering: Example of community middleware extending or customizing DeepAgents behavior.
- AWS: Infrastructure context for scaling evaluation of DeepAgents.
Newsletter Mentions (10)
“Harrison Chase clarifies that DeepAgents isn’t runtime lock-in—being OS-based, you can run it anywhere, whether in SuperQode with a different runtime, LangGraph, Temporal, or any other platform.”
The item frames DeepAgents as a portable agent system rather than a platform-locked one.
“Harrison Chase observes the agent industry pivoting from frameworks like LangChain, AI SDK, and LlamaIndex to full-fledged harnesses such as DeepAgents, Claude Agent SDK, and EVE—with DeepAgents predating EVE by about ten months.”
#2 𝕏 Harrison Chase observes the agent industry pivoting from frameworks like LangChain, AI SDK, and LlamaIndex to full-fledged harnesses such as DeepAgents, Claude Agent SDK, and EVE—with DeepAgents predating EVE by about ten months.
“Harrison Chase shows how to build a live voice agent by offloading complex reasoning to DeepAgents and using Gemini Live for natural, low-latency interactions.”
Harrison Chase shows how to build a live voice agent by offloading complex reasoning to DeepAgents and using Gemini Live for natural, low-latency interactions. #15 📝 Claude Code Blog Getting started with loops - A tutorial-style post introducing loops in Claude Code, aimed at helping developers get started using loop constructs and workflows.
“#8 𝕏 Harrison Chase introduced dynamic subagents in Deepagents, letting you programmatically spin up subagents and showcasing six distinct use cases.”
#8 𝕏 Harrison Chase introduced dynamic subagents in Deepagents, letting you programmatically spin up subagents and showcasing six distinct use cases.
“Harrison Chase highlights a nearly 10-hour agentic AI course covering LangChain, LangGraph, RAG, deepagents and guardrails.”
#3 𝕏 Harrison Chase highlights a nearly 10-hour agentic AI course covering LangChain, LangGraph, RAG, deepagents and guardrails. He’s also asking for other strong Lang* resources for learners.
“Harrison Chase breaks down how to evaluate DeepAgents at scale on AWS with LangSmith, covering concrete datapoint and evaluator design methods for longer-horizon agents.”
#2 𝕏 Harrison Chase breaks down how to evaluate DeepAgents at scale on AWS with LangSmith, covering concrete datapoint and evaluator design methods for longer-horizon agents.
“#19 𝕏 Harrison Chase predicts that by 2026 closed-model costs will be prohibitively high and he’s optimizing deepagents for peak performance on OSS models.”
#19 𝕏 Harrison Chase predicts that by 2026 closed-model costs will be prohibitively high and he’s optimizing deepagents for peak performance on OSS models. #20 in Peter Yang showcases five new Google Labs AI products—Pomelli for marketing, Stitch for design, Genie for 3D worlds, Flow for video creation, and NotebookLM for research synthesis.
“#1 𝕏 Harrison Chase launched Deepagents—an open-source Claude Agent SDK available at github.com/langchain-ai/deepagents for building and running Claude-powered agents.”
GenAI PM Daily April 13, 2026 GenAI PM Daily 🎧 Listen to this brief 3 min listen Today's top 14 insights for PM Builders, ranked by relevance from X, Blogs, and YouTube. Deepagents Releases Open Source Claude Agent SDK #1 𝕏 Harrison Chase launched Deepagents—an open-source Claude Agent SDK available at github.com/langchain-ai/deepagents for building and running Claude-powered agents.
“Harrison Chase notes that community middleware—like “langchain-task-steering”—is popping up for customizing agents and deepagents, and invites anyone with middleware to contribute by reaching out to Sydney.”
#25 𝕏 Harrison Chase notes that community middleware—like “langchain-task-steering”—is popping up for customizing agents and deepagents, and invites anyone with middleware to contribute by reaching out to Sydney.
“No-code AI agent builder goes GA : Harrison Chase @hwchase17 announced LangSmith Agent Builder is now generally available, featuring **deepagents**, **memory**, **MCPS/skills/subagents**, **triggers** for autonomous workflows, and an **agent inbox** for human-in-the-loop review.”
AI Tools & Applications. Agentic file exploration vs. RAG : LlamaIndex @llama_index shared results from an experiment comparing an **fs-explorer agent** against **hybrid RAG**, highlighting when agent-centric file search offers advantages over traditional vector retrieval. No-code AI agent builder goes GA : Harrison Chase @hwchase17 announced LangSmith Agent Builder is now generally available, featuring **deepagents**, **memory**, **MCPS/skills/subagents**, **triggers** for autonomous workflows, and an **agent inbox** for human-in-the-loop review.
Related
Anthropic is the company behind Claude and Claude Code. The newsletter covers its new Reflection dashboard and an enterprise deployment of Claude in industrial workflows.
Anthropic’s assistant and coding tool, discussed here in both the Reflection dashboard and a physical-AI deployment at UST. The newsletter highlights its usage analytics, workflow suggestions, and enterprise integration.
LlamaIndex is referenced as a company/brand running ParseBench against GPT-5.6. The note highlights its use in evaluating document parsing performance.
Founder and/or public builder associated with LangSmith, LangChain, and LLM knowledge tooling. He is mentioned launching LangSmith and hosting an LLM Wiki Webinar.
An AI infrastructure company known for building tools for LLM apps and agents. In this newsletter, it is associated with DeepAgents and open-source coding infrastructure.
A cloud platform for agent orchestration, observability, sandboxes, and deployments. It is presented as integrated with many LangChain models and designed for recursive improvement loops.
An SDK for building Claude-based agents and workflows. It is cited as one of the newer harness-style tools replacing older frameworks.
A developer framework for building AI-enabled applications, mentioned as part of the prior generation of agent tooling. It is contrasted with newer end-to-end harnesses.
Cloud platform provider appearing in multiple enterprise and agent infrastructure contexts. In this newsletter it is associated with Claude Desktop availability and AgentCore Payments.
Community middleware example for customizing agent behavior and steering tasks in agent frameworks.
Stay updated on deepagents
Get curated AI PM insights delivered daily — covering this and 1,000+ other sources.
Subscribe Free