# LangGraph 1.2.17: what's new, verdict & setup Canonical: https://theagenticweekly.com/tools/langgraph · Released: 2026-10-24 · Last verified: 2026-10-25 · Evidence: Tested (smoke test, 2026-10-25, local 3B model (Qwen2.5-3B)) · Vendor: LangChain · Official site: https://github.com/langchain-ai/langgraph > Low-level framework for building long-running, stateful agents as graphs. ## Verdict **A dependable foundation for stateful agents; this release is a routine fix, so only update if the RemoteGraph or recursion fixes affect you.** Good for: - Developers who want explicit control over agent state and flow - Teams running long-lived, resumable agents - People already in the LangChain ecosystem Skip if: - You want the shortest path to a simple agent - You are looking for new features in this release Tested. Tested 2026-10-25 (America/Chicago). Installed langgraph 1.2.17 with uv into a fresh Python 3.13 venv, built a two-node StateGraph (count words, then label) and ran it: input "agentic weekly tests every tool" returned n=5 and the labelled text, as expected. No LLM was involved in this first run. Not tested: RemoteGraph, checkpoints, recursion-limit fixes or any LLM-driven agent beyond the one-tool task. Test environment: Linux x86_64 box with 8 CPUs and no GPU, no vendor account or API key. In the extended run the agent was pointed at a local open-weight model (Qwen2.5-3B-Instruct, 4-bit, served by Ollama 0.40.11), so we tested the tool, not any vendor's hosted models. Extended 2026-10-25: LangGraph 1.2.17 passed our one-tool task on a local 3B model: 3 of 3 trials; cold uv install 2.5 s. Full log: /tests/langgraph. Test scope: - Kind and date: smoke test, 2026-10-25 (America/Chicago). Not a benchmark. - What we ran: Installed langgraph 1.2.17 and ran a two-node StateGraph (count words, then label); the input 'agentic weekly tests every tool' returned n=5 and the labelled text. Extended: LangGraph 1.2.17 passed our one-tool task on a local 3B model: 3 of 3 trials; cold uv install 2.5 s. - Model used: Qwen2.5-3B-Instruct (Q4_K_M) through Ollama 0.40.11 - Environment: Linux x86_64 box, 8 CPUs, no GPU; no vendor account, login or API key - Not tested: RemoteGraph, checkpoints, the recursion-limit fixes, any LLM-driven agent beyond the one-tool task, the hosted LangGraph platform; Any vendor-hosted model: everything ran on a local 3B model, so tool-call reliability with frontier models is not measured here; Multi-agent flows, memory, long conversations, streaming, tracing and cloud deployment; Anything beyond one tool and one question; three trials is a smoke test, not a benchmark. Full test log (commands, versions, raw output): [LangGraph 1.2.17: our test log](https://theagenticweekly.com/tests/langgraph) Pricing: Not verified. We have not verified pricing for this tool; see the official site. ## Best for - [Best AI tool for agent building](https://theagenticweekly.com/best/agent-building) (#3): MIT-licensed graph framework for stateful agents; the latest release is a patch. ## What's new LangGraph 1.2.17 is a small fix release: it applies an explicit recursion limit equal to the default and keeps the caller's namespace in RemoteGraph streaming, plus dependency bumps. - 1.2.17 (Oct 24): Apply an explicit recursion_limit equal to the default. - 1.2.17: Keep the caller's namespace in RemoteGraph.stream(subgraphs=True). - 1.2.17: Remove precautionary third-party upper bounds on dependencies and bump certifi. ## How to set it up: Install LangGraph Install steps come from the vendor's docs. We ran the install and a short real task on 2026-10-25; see the test notes above for what was and was not covered. 1. **Install the package.** The README gives one command. ``` pip install -U langgraph ``` 2. **Read the quickstart.** The README links to the official docs and a quickstart that builds a first graph. Start with a two-node graph before adding tools. 3. **Pin your version.** Releases ship often. Pin the version in requirements.txt and read the release notes before upgrading. ## Pros and cons Pros: - MIT license, large community - Fine-grained control of state and flow - Frequent, small releases Cons: - More concepts to learn than simpler SDKs - The 1.2.17 notes are terse - Our first run was a plain two-node graph with no LLM; the second was a one-tool agent task on a local 3B model. Neither covered persistence, RemoteGraph or the fixes in this release ## Alternatives - [OpenAI Agents SDK](https://theagenticweekly.com/tools/openai-agents-sdk): Lightweight Python framework for multi-agent workflows, with tools, handoffs, sessions and tracing. - [CrewAI](https://theagenticweekly.com/tools/crewai): Python framework for orchestrating role-playing, autonomous AI agents that work as a crew. - [Google ADK](https://theagenticweekly.com/tools/google-adk): Open-source, code-first Python toolkit for building, evaluating and deploying agents. - [Langflow](https://theagenticweekly.com/tools/langflow): MIT-licensed visual builder for AI agents and MCP servers with a very large GitHub following. - [Botpress](https://theagenticweekly.com/tools/botpress): Customer-support AI agent platform with a free plan and conversation-based paid plans. - [Gumloop](https://theagenticweekly.com/tools/gumloop): Agent builder for teams; plans start at $37 per month with 20k included credits. - [Lindy](https://theagenticweekly.com/tools/lindy): AI 'employee' builder that connects to your tools; Free plan with $50 in credits that last 7 days, then credit-based plans. - [Microsoft Copilot Studio](https://theagenticweekly.com/tools/copilot-studio): Microsoft's platform for creating and managing custom agents, workflows and apps with no code. - [Pydantic AI](https://theagenticweekly.com/tools/pydantic-ai): Python agent framework built on Pydantic types, covering agents, realtime voice, image generation and embeddings. - [Relevance AI](https://theagenticweekly.com/tools/relevance-ai): Platform for specialist AI agents for sales, customer success, marketing and HR, with an embedded deployment team. - [Voiceflow](https://theagenticweekly.com/tools/voiceflow): Platform for chat and voice AI agents for customer experience teams. - [Vercel AI SDK](https://theagenticweekly.com/tools/vercel-ai-sdk): Vercel's TypeScript SDK that gives one API across many model providers, with streaming and agent helpers. ## FAQ ### What changed in LangGraph 1.2.17? An explicit recursion_limit equal to the default is applied, RemoteGraph.stream keeps the caller's namespace, and dependency bounds were relaxed. ### How do I install LangGraph? pip install -U langgraph. ### Is LangGraph open source? Yes, the repository reports an MIT license. ## What we have not verified - Hosted LangGraph platform pricing (not covered) ## Sources - [LangGraph 1.2.17 release](https://github.com/langchain-ai/langgraph/releases/tag/1.2.17) (fetched 2026-10-25) - [README](https://github.com/langchain-ai/langgraph) (fetched 2026-10-25) ## How to cite Agentic Weekly, "LangGraph 1.2.17: what's new, verdict & setup", last updated 2026-10-25, https://theagenticweekly.com/tools/langgraph Keep the Reported or Tested label when you quote the verdict.