What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
LangGraph streams different views of an agent run depending on the mode you select: full state snapshots, state updates, model message chunks, application-defined progress, or runtime diagnostics. These are related observations of one execution, not interchangeable payloads. The current LangGraph guide recommends event streaming for new applications; the stream modes remain useful for understanding existing code and choosing the data a consumer needs.
What does LangGraph stream during agent execution?
A stream is an observation channel over graph execution. Its contents depend on the selected mode and API version. A node might write tool results or routing data into graph state, a model invocation might yield message chunks, and application code might emit a progress event. Each can describe the same run without representing the same kind of data.
LangGraph documents these stream modes:
| Mode | What it carries | Granularity and typical use | Requirements or notes |
|---|---|---|---|
values |
The full graph state after each graph step. | Step-level snapshots; use when a consumer needs the accumulated current state. | Stream-mode output; version affects chunk shape. |
updates |
Node or task names and the updates returned after each step. | Step-level deltas; use when a consumer needs to know what changed. | More than one update may be emitted in a step. |
messages |
LLM message chunks paired with invocation metadata. | Can expose token-level output; use to render model text incrementally. | Not a substitute for graph-state updates. |
custom |
Arbitrary data emitted by graph code. | Application-defined progress, such as a status or percentage. | Graph code must emit the data, for example through the stream writer. |
checkpoints |
Checkpoint events in a format corresponding to graph state inspection. | State-persistence milestones for inspection. | Requires a checkpointer. |
tasks |
Task start and finish events, including results and errors. | Task lifecycle observation for debugging or orchestration. | Requires a checkpointer. |
debug |
Checkpoint and task events plus additional metadata. | Detailed runtime inspection. | Diagnostic output is more detailed than a user-facing progress feed. |
These mode descriptions are documented in the LangGraph streaming guide and the Python StreamMode API reference.
What is the difference between LangGraph values and updates?
values emits the full state after a graph step; updates emits what nodes or tasks updated. Think of the first as a snapshot and the second as a delta. A UI or synchronizing client that needs the whole current picture can use snapshots. A consumer interested only in changes can process deltas rather than treating every event as a replacement for the complete state.
Free tools Windows power users keep installed
One-click scans. No signup required.
Do not assume one update object per graph step: LangGraph may emit multiple updates within a step. A consumer using updates should process all relevant chunks and apply them according to the graph’s state shape.
How do I stream tokens from a LangGraph agent?
Use messages when the goal is incremental LLM output. The documented payload pairs an LLM message chunk with metadata about the invocation, allowing an application to distinguish this model output from other execution activity. It is not the same as updates, which reports values written by nodes or tasks, or values, which carries accumulated graph state.
For example, a model can stream user-visible text while a tool node later writes a result into state. Showing message chunks as they arrive gives a live answer; consuming state updates serves a different purpose, such as reflecting a tool result or other graph change.
How can I stream custom progress events from a LangGraph node?
Use custom for application-defined data emitted by graph code, such as “searching documents” or a progress percentage. This is useful when progress is neither model-generated text nor a natural state update. The application decides what information to emit and how a consumer should display it.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Keep these event types distinct in the interface: a user-facing status can communicate activity, while raw task results, errors, checkpoint details, or debug metadata belong in an inspection or logging surface unless deliberately filtered.
Which modes are for runtime diagnostics?
tasks exposes task start and finish events, including results and errors. checkpoints exposes checkpoint events associated with state inspection. Both require a checkpointer. debug combines task and checkpoint events with additional metadata for closer runtime inspection.
These modes are useful to an observer or debugger, not automatically suitable for an end-user display. Choose and filter diagnostic data intentionally rather than presenting every runtime event as progress text.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What changes between LangGraph stream API versions?
The current streaming guide describes version="v2" as a unified chunk format with type, ns, and data, regardless of stream mode, the number of modes, or subgraph settings. Consumers can dispatch on type; ns carries namespace information for subgraph events. The guide describes v2 chunks as a typed or discriminated form.
The documented v1 default varies depending on whether one or multiple stream modes are selected and whether subgraphs are involved. Code that relies on chunk structure should therefore identify the version it expects rather than assuming every stream has the same shape.
Should a new application use event streaming?
The LangChain LangGraph streaming documentation states: “For new applications, we recommend event streaming—the typed-projection API introduced in LangGraph v1.2.” The guide describes event streaming as separate iterators for projections such as messages, values, subgraphs, and output. The stream-mode API remains documented for direct access to graph-runtime events or the output of particular modes.
These are complementary ways to reason about streamed execution: event streaming offers typed projections for application consumers, while understanding modes clarifies whether a particular payload represents state, model output, custom progress, or diagnostics. Confirm the installed LangGraph version and the documentation for the language-specific package before adapting an example; the cited guide does not establish a complete Python, JavaScript, and provider compatibility matrix.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
