# From Slack: Parallel Nodes

**URL:** <https://forum.langchain.com/t/from-slack-parallel-nodes/2208>\
**Category:** LangGraph\
**Tags:** python-help\
**Created:** [November 14, 2025, 1:15pm UTC](https://forum.langchain.com/t/from-slack-parallel-nodes/2208 "2025-11-14T13:15:38Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![MchoneDev](https://yyz1.discourse-cdn.com/flex007/user_avatar/forum.langchain.com/mchonedev/32/806_2.png) [@MchoneDev](https://forum.langchain.com/u/MchoneDev)\
**Post date:** [November 14, 2025, 1:15pm UTC](https://forum.langchain.com/t/from-slack-parallel-nodes/2208/1 "2025-11-14T13:15:38Z")

</div>

**Jithin Shaji** asks:

In langgraph how do you make a certain router output start 2 nodes in parallel?

---

<div class="post-metadata">

**Author:** ![MchoneDev](https://yyz1.discourse-cdn.com/flex007/user_avatar/forum.langchain.com/mchonedev/32/806_2.png) [@MchoneDev](https://forum.langchain.com/u/MchoneDev)\
**Post date:** [November 14, 2025, 1:18pm UTC](https://forum.langchain.com/t/from-slack-parallel-nodes/2208/2 "2025-11-14T13:18:02Z")

</div>

In order to have a node output parallel lines of work, you can either use the `send` api in order to output dynamic next steps: [Graph API overview - Docs by LangChain](https://docs.langchain.com/oss/python/langgraph/graph-api#send)

Another alternative is that if this parallel work is determined ahead of time, you can setup multiple edges coming from the same node to the parallel destinations.

> A node can have multiple outgoing edges. If a node has multiple outgoing edges, all of those destination nodes will be executed in parallel as a part of the next superstep.

> **[Graph API overview - Docs by LangChain](https://docs.langchain.com/oss/python/langgraph/graph-api#edges)**

---

<div class="post-metadata">

**Author:** ![jitdarkfighter](https://yyz1.discourse-cdn.com/flex007/user_avatar/forum.langchain.com/jitdarkfighter/32/1633_2.png) [@jitdarkfighter](https://forum.langchain.com/u/jitdarkfighter)\
**Post date:** [November 15, 2025, 3:50am UTC](https://forum.langchain.com/t/from-slack-parallel-nodes/2208/4 "2025-11-15T03:50:02Z")

</div>

Yeah, Set up a dummy node that returns the state, and then this dummy node will route to the 2 required nodes assume (A and B). So now A and B will run parallel once and if routed to A again, only A will run and not B.

Also does langchain need some barriers for parallel execution or does it handle it internally. For example A can be said to be a graph on its own and B redirects to END.