Studio troubleshooting ↗

    pip install -U langgraph-cli>=0.2.6
    langgraph dev --tunnel
    ```
  <span class="tab-end"></span>

  <span class="tab-start" data-tab-title="JS"></span>
```shell
    # Requires @langchain/langgraph-cli>=0.0.26
    npx @langchain/langgraph-cli dev --tunnel
    ```
  <span class="tab-end"></span>
<span class="tab-group-end"></span>

The command outputs a tunnel URL. To connect Studio:

1. Copy the tunnel URL (e.g., `https://hamilton-praise-heart-costumes.trycloudflare.com`)
2. Open Studio at `https://smith.langchain.com/studio/`
3. Click **Connect to a local server**
4. Paste the tunnel URL and add it to **Allowed Origins**
5. Click **Connect**

This manual step is required for security - Studio requires explicit user confirmation before connecting to external URLs.

<span class="callout-start" data-callout-type="note"></span>
  Cloudflare tunnels can be unreliable and may intermittently disconnect.
<span class="callout-end"></span>

### Solution 2: Use Chromium browser

Chrome and other Chromium browsers allow HTTP on localhost. Use `langgraph dev` without additional configuration.

## Chrome connection issues

Starting with Chrome version 142, you may experience "Failed to initialize Studio" errors with "TypeError: Failed to fetch" when trying to connect [LangSmith Studio](/langsmith/studio) to your local development server via [`langgraph dev`](/langsmith/cli). This occurs even when the API server at `http://127.0.0.1:2024/docs` loads successfully.

**Root Cause:** Chrome 142 fully enforces the Private Network Access (PNA) specification with no fallback, which blocks HTTPS sites (like `https://smith.langchain.com`) from accessing HTTP localhost servers by default.

### Symptoms

* Running `langgraph dev` starts the server successfully.
* Navigating to `http://127.0.0.1:2024/docs` shows the API documentation correctly.
* LangSmith Studio at `https://smith.langchain.com` shows: "Failed to initialize Studio - Please verify if the API server is running or accessible from the browser. TypeError: Failed to fetch".
* Browser console shows errors like: `Permission was denied for this request to access the 'unknown' address space`.

### Solution: Allow local network access in Chrome

1. Open LangSmith Studio at `https://smith.langchain.com` in Chrome.
2. Click the **lock icon** (or site information icon) to the left of the address bar.
3. Look for the **"Local network access"** option in the dropdown.
4. Change the setting from **"Ask (default)"** or **"Block"** to **"Allow"**.
5. Reload the page.

Studio should now connect to your local development server successfully.

### Additional troubleshooting

**Check for browser extension conflicts**

Browser extensions (especially Ollama Chrome extension or AI model extensions) can interfere with localhost connections:

1. Disable all browser extensions temporarily.
2. Restart Chrome.
3. Try connecting to Studio again.
4. If it works, re-enable extensions one by one to identify the culprit.

**Verify dependencies are up to date**

```shell
pip install -U "langgraph-cli[inmem]"

    pip install -U langgraph-cli>=0.2.6
    langgraph dev --tunnel
    ```
  <span class="tab-end"></span>

  <span class="tab-start" data-tab-title="JS"></span>
```shell
    # Requires @langchain/langgraph-cli>=0.0.26
    npx @langchain/langgraph-cli dev --tunnel
    ```
  <span class="tab-end"></span>
<span class="tab-group-end"></span>

The command outputs a tunnel URL. To connect Studio:

1. Copy the tunnel URL (e.g., `https://hamilton-praise-heart-costumes.trycloudflare.com`)
2. Open Studio at `https://smith.langchain.com/studio/`
3. Click **Connect to a local server**
4. Paste the tunnel URL and add it to **Allowed Origins**
5. Click **Connect**

This manual step is required for security - Studio requires explicit user confirmation before connecting to external URLs.

## Graph edge issues

Undefined conditional edges may show unexpected connections in your graph. This is
because without proper definition, Studio assumes the conditional edge could access all other nodes. To address this, explicitly define the routing paths using one of these methods:

### Solution 1: Path map

Define a mapping between router outputs and target nodes:

<span class="tab-group-start"></span>
  <span class="tab-start" data-tab-title="Python"></span>
```python
    graph.add_conditional_edges("node_a", routing_function, {True: "node_b", False: "node_c"})
    ```
  <span class="tab-end"></span>

  <span class="tab-start" data-tab-title="Javascript"></span>
```ts
    graph.addConditionalEdges("node_a", routingFunction, { true: "node_b", false: "node_c" });
    ```
  <span class="tab-end"></span>
<span class="tab-group-end"></span>

<a id="python" />

### Solution 2: Router type definition

Specify possible routing destinations using Python's `Literal` type:

```python
def routing_function(state: GraphState) -> Literal["node_b","node_c"]:
    if state['some_condition'] == True:
        return "node_b"
    else:
        return "node_c"