Langflow release notes | Langflow Documentation (original) (raw)

This page summarizes significant changes to Langflow in each release. For all changes, see the Changelog.

Due to strict SemVer requirements, Langflow Desktop can have different patch versions than the core Langflow OSS Python package, but the major and minor versions are aligned.

warning

Whenever possible, the Langflow team recommends installing new Langflow versions in a new virtual environment or VM before upgrading your primary installation. This allows you to import flows from your existing installation and test them in the new version without disrupting your existing installation. In the event of breaking changes or bugs, your existing installation is preserved in a stable state. If you are upgrading Windows Desktop to 1.6.0, don't auto-upgrade with the in-app Update button. Instead, follow the instructions in Known issue: Don't auto-upgrade Windows Desktop.

To avoid the impact of potential breaking changes and test new versions, the Langflow team recommends the following upgrade process:

1. Recommended: Export your projects to create backups of your flows:
   `
   _10
   curl -X GET \
   _10
   "$LANGFLOW_SERVER_URL/api/v1/projects/download/$PROJECT_ID" \
   _10
   -H "accept: application/json" \
   _10
   -H "x-api-key: $LANGFLOW_API_KEY"

`
To export flows from the visual editor, see Import and export flows. 2. Install the new version:

• Langflow OSS Python package: Install the new version in a new virtual environment. For instructions, see Install and run the Langflow OSS Python package.
• Langflow Docker image: Run the new image in a separate container.
• Langflow Desktop: To upgrade in place, open Langflow Desktop, and then click Upgrade Available in the Langflow header. If you want to isolate the new version, you must install Langflow Desktop on a separate physical or virtual machine, and then import your flows to the new installation.

1. Import your flows to test them in the new version, upgrading components as needed.
   When upgrading components, you can use the Create backup flow before updating option if you didn't previously export your flows.
2. If you installed the new version in isolation, upgrade your primary installation after testing the new version.
   If you made changes to your flows in the isolated installation, you might want to export and import those flows back to your upgraded primary installation so you don't have to repeat the component upgrade process.

Highlights of this release include the following changes. For all changes, see the Changelog.

Breaking changes​

• langflow-base dependency structure refactored
  The langflow-base package now uses granular optional dependency groups. As a result, many dependencies that were previously included in the langflow-base installation were moved to optional extras.
  If you installed Langflow with uv pip install langflow, this isn't a breaking change. Installing langflow automatically installs langflow-base[complete], which includes all optional dependencies and maintains the same functionality as before.
  However, if you installed Langflow with uv pip install langflow-base without specifying extra dependencies, this is a breaking change. Some dependencies that were previously included by default are now available only through optional extras. Therefore, installing langflow-base directly only installs the core base dependencies.
  If you installed langflow-base, there are two ways to resolve dependency errors that result from this breaking change:

  • If you need the full set of dependencies, you must install langflow-base with the complete extra:
    `
    _10
    uv pip install "langflow-base[complete]"

  `

  • If you need specific dependencies, you must install langflow-base with those optional dependency groups. For example:
    `
    _10
    uv pip install "langflow-base[postgresql,openai,chroma]"

  `

For more information about available optional dependency groups, see Install optional dependency groups for langflow-base.

• Docker image does not include uv or uvx
  tip
  Starting with Langflow 1.8.1, the official Docker images include uv and uvx again. If you're using Langflow 1.8.0, follow the steps in this release note to add uv and uvx in a derived image.
  In Langflow 1.8.0, the Langflow Docker image does not include uv or uvx in the runtime image. This means that MCP server configurations, including the default Langflow MCP server, that rely on commands like uvx mcp-proxy will fail inside the container with a command not found error.
  If you use MCP from within a Langflow 1.8.0 Docker image, you must install uv in an image derived from the official langflowai/langflow image.
  To install uv and uvx in a derived image based on the official langflowai/langflow image, do the following:

  1. Create a Dockerfile in your project.
  2. Use the official Langflow image as the base and install uv into the existing virtual environment:
     `
     _10
     FROM langflowai/langflow:1.8.0
     _10
     _10
     USER root
     _10
     RUN /app/.venv/bin/pip install uv
     _10
     USER user

  3. Build the image.
  _10
  docker build -t YOUR_IMAGE_NAME .

  ReplaceYOUR_IMAGE_NAMEwith your Docker image name. 4. Run the new image:
  _10
  docker run -p 7860:7860 YOUR_IMAGE_NAME

  `

For more information, see Upgrade the Langflow Docker image.

• Disabled tool calling for Gemini 3 preview models
  Tool calling is temporarily disabled for the following Google Gemini 3 preview models:gemini-3.1-pro-preview, gemini-3-pro-preview, gemini-3-flash-preview, and gemini-3-pro-image-preview.
• Docling installation behavior differs between Desktop and OSS
  In Langflow 1.8.0, Docling behavior depends on how Langflow is installed.
  For Langflow Desktop installations, you need to set LANGFLOW_DOCLING=True in your .env file to enable Docling dependency installation. For more information, see Set environment variables for Langflow Desktop.
  For Langflow OSS installations with uv pip install langflow, Docling is included with the Langflow package.
  For Langflow OSS installations with uv pip install langflow-base, install Docling explicitly as an extra with uv pip install "langflow-base[docling]".

New features and enhancements​

• Global model provider configuration
  Model providers for language models, embedding models, and agents are now configured globally in the Model providers pane, instead of within individual components. For more information, see the Language Model component.
• Component inspection panel
  The component inspection panel replaces the component header menu for managing component parameters and settings. For more information, see Component inspection panel.
• Developer API: /workflow synchronous endpoints (Beta)
  The Developer API is part of a larger effort to improve Langflow's APIs with enhanced capabilities and better developer experience. The Developer API now includes /v2/workflow endpoints for executing flows with enhanced error handling, timeout protection, and structured responses. The synchronous execution endpoint is available at POST /api/v2/workflows. For more information, see Workflow API (Beta).
• Traces and trace view
  Langflow now records execution traces for flows and components. View your traces in the Trace Details pane, and inspect span trees, latencies, and errors. For more information, see Traces.
• Knowledge bases
  Knowledge bases let you organize documents and other reference data into reusable vector databases that can be attached to multiple flows. This makes it easier to centralize domain knowledge and reuse the same data across agents and retrieval workflows. For more information, see Manage vector data.
• Mustache templating support for Prompt Template component
  The Prompt Template component now supports Mustache templating syntax. Mustache templating eliminates the need to escape curly braces when including JSON structures in your prompts. For more information, see Prompt Template.
• More configuration options for JWT-based session authentication
  Langflow 1.8 offers additional configuration options for JWT algorithms, including support for RS256/RS512 algorithms, configurable keys, and token lifetimes. For more information, see JWT authentication.
• Global variables in MCP server headers
  You can now use global variables in MCP server header values to securely store and reference sensitive values. For more information, see Use global variables in MCP server headers.
• Pass environment variables to flows in API headers and CLI
  The ability to pass environment variables in HTTP headers (previously available for the /responses endpoint) is now also available for the /run endpoint.
• Guardrails component
  The Guardrails component validates input text against security and safety guardrails by using a connected language model to check for content such as PII, tokens/passwords, or offensive content. For more information, see Guardrails.
• Token usage tracking for OpenAI Responses API
  The OpenAI Responses API endpoint now tracks and returns token usage statistics when your flow uses language model APIs that provide token usage information. For more information, see Token usage tracking.
• Docker AMD vs ARM image sizes
  Langflow 1.8.0 addresses the AMD vs ARM Docker image size gap. We reconfigured our Python dependencies to use CPU-only PyTorch wheels through uv sources, which removes large CUDA-related dependencies from the AMD64 images. With this change, both AMD64 and ARM64 images are now smaller than 2 GB.
• New Agentics bundle
  Uses LLMs to transform tabular data, including mapping, reducing, and generating DataFrame rows based on a defined schema.
• New LiteLLM bundle
  Connects to models through a LiteLLM proxy so you can route requests to multiple LLM providers and switch providers without changing flow credentials.
• New Openlayer observability integration
  Configures Langflow to send tracing data to Openlayer for analysis, monitoring, and evaluation of your flow executions.

Version yanked

Version 1.7.0 was yanked due to a critical bug. Version 1.7.0 has been replaced with version 1.7.1, which includes a fix for this issue.

Highlights of this release include the following changes. For all changes, see the Changelog.

Known issue: Data not accessible when upgrading to version 1.7.0​

A critical issue was identified during the upgrade process to version 1.7.0.

Flows, projects, and global variables are not deleted or corrupted. The data still exists, but version 1.7.0 cannot find it due to a path change in how flows are located. All Langflow versions upgrading to 1.7.0 are affected.

Don't upgrade to Langflow version 1.7.0. Instead, upgrade directly to version 1.7.1, which includes a fix for this bug.

If you installed version 1.7.0 before the fix was released, follow these steps to recover your flows:

1. Revert Langflow to version 1.6.9:
   `
   _10
   uv pip install langflow==1.6.9

2. Verify that your flows, projects, and global variables are accessible. 3. Upgrade directly to version 1.7.1, which includes the fix for this issue:
_10
uv pip install langflow==1.7.1

`

New features and enhancements​

• Support for streamable HTTP transport for MCP clients and servers
  Langflow now supports streamable HTTP transport for both MCP clients and servers. When using Langflow as an MCP client, you can connect to MCP servers with streamable HTTP transport. When using Langflow as an MCP server, clients can connect using streamable HTTP transport. SSE transport is still supported as a fallback for backwards compatibility.
• Webhook authentication
  Added the LANGFLOW_WEBHOOK_AUTH_ENABLE environment variable for authenticating requests to the /webhook endpoint. When LANGFLOW_WEBHOOK_AUTH_ENABLE=TRUE, webhook endpoints require API key authentication and validate that the authenticated user owns the flow being executed. When FALSE, no Langflow API key is required and all requests to the webhook endpoint are treated as being sent by the flow owner. For more information, see Trigger flows with webhooks.
• Configurable API key validation
  Added the LANGFLOW_API_KEY_SOURCE environment variable to control how Langflow validates API keys. When set to db, Langflow validates API keys against keys stored in the database. When set to env, Langflow validates API keys against the LANGFLOW_API_KEY environment variable. For more information, see API keys and authentication.
• SSRF protection
  Added SSRF (Server-Side Request Forgery) protection to the API Request component. HTTP redirects are disabled by default to prevent SSRF bypass attacks. To enable SSRF protection, set LANGFLOW_SSRF_PROTECTION_ENABLED=TRUE. Configure allowed hosts with LANGFLOW_SSRF_ALLOWED_HOSTS. Flows that relied on automatic redirects will need to enable it manually.
• Email registration in Langflow Desktop
  Langflow Desktop now includes an email registration screen at startup. The registered email address helps track user registrations and understand the user base. For more information, see Telemetry.
• Changes to read/write file components
  The Save File component was renamed to Write File, and it can now save to S3 and Google Drive. The File component was renamed to Read File, and it can now read from AWS S3 and Google Drive. Both components support Tool Mode.
• New integrations, bundles, and components:
  New filter operator for DataFrame Operations component
  The DataFrame Operations component now includes a not contains filter operator. Use it to clean data by extracting only records that don't contain specific values. For example, you can filter out invalid email addresses that don't contain @.
  New JSON operations for Data Operations component
  The Data Operations component now includes two operations for advanced JSON data manipulation. The Path Selection operation extracts values from nested JSON structures, and the JQ Expression operation uses the jq query language to perform advanced JSON filtering, projections, and transformations.
  New Smart Router component
  New Mock Data component
  New Dynamic Create Data component
  New ALTK bundle
  New CometAPI bundle
  New CUGA bundle
  The LLM Router component is now called the LLM Selector component.
  The Web Search component now consolidates Web Search, News Search, and RSS Reader into a single component with tabs for different search modes. You can search the web using DuckDuckGo, search Google News, or read RSS feeds—all from one component. The separate News Search and RSS Reader components have been removed.

Highlights of this release include the following changes. For all changes, see the Changelog.

Known issue, potential security vulnerability: .env file not loaded in versions 1.6.0 through 1.6.3​

Langflow versions 1.6.0 through 1.6.3 have a critical bug where environment variables from .env files aren't read. This affects all deployments using environment variables for configuration, including security settings.

Potential security vulnerability

If your .env file includes AUTO_LOGIN=false, upgrading to the impacted versions causes Langflow to fall back to default settings, potentially giving all users superuser access immediately upon upgrade. Additionally, database credentials, API keys, and other sensitive configurations can't be loaded from .env files.

Don't upgrade to any Langflow version from 1.6.0 through 1.6.3 if you use .env files for configuration. Instead, upgrade to 1.6.4, which includes a fix for this bug.

Known issue: Don't auto-upgrade Windows Desktop​

warning

Windows users of Langflow Desktop should not use the in-app update feature to upgrade to Langflow version 1.6.0. Only Windows Desktop users upgrading to Langflow version 1.6.0 are affected.

The Update button in Langflow Desktop will not work for Windows users, and may result in data loss.

Instead, download a fresh installation from the Langflow website when a new version is available.

Follow the instructions below to minimize the risk of losing flows.

These instructions assume the default path for the Langflow database of C:\Users\YOUR_USERNAME\AppData\Roaming\com.langflow\data\database.db or C:\Users\%YOUR_USERNAME%\AppData\Roaming\com.langflow\data\database.db. Replace YOUR_USERNAME with your Windows username.

1. Ensure you have Administrator privileges.
2. Ensure you have enough disk space for a second installation of Langflow.
3. Close Langflow, and ensure no Langflow process is running in Task Manager.
4. The Langflow 1.6.0 installer automatically performs a database backup in a later step, but an additional manual backup provides additional redundancy and is recommended. To manually back up your Langflow database file, do the following:
   1. In Windows Explorer, navigate to C:\Users\YOUR_USERNAME\AppData\Roaming\com.langflow\data\. Replace YOUR_USERNAME with your Windows username.
   2. Copy the database.db at this location, and paste it to a safe location.
5. Download the Windows installer from the Langflow website.
6. Run the Windows installer as an Administrator. To run the installer as an Administrator, right-click the executable and select Run as administrator.
7. Follow the Windows installer's guided steps. The Langflow 1.6.0 installer automatically performs a database backup. These steps install Langflow from scratch, and result in two Langflow installations: the previously installed version, and version 1.6.0. This is expected behavior.
8. Start version 1.6.0 of Langflow, and confirm your flows behave as expected.
9. If flows are missing, restore your flows from your manual backup by doing the following:
   1. Close Langflow.
   2. Navigate to your backup location, and copy the database.db file.
   3. Replace the database file in the new installation's data directory.
   4. Start Langflow, and confirm your flows behave as expected.
   5. After confirmation, uninstall the previous version of Langflow, and keep version 1.6.0.

Breaking changes​

• Authentication enforced for Langflow API requests by default
  In Langflow version 1.6, LANGFLOW_AUTO_LOGIN=True and LANGFLOW_SKIP_AUTH_AUTO_LOGIN=False by default. This enforces authentication for Langflow API requests while still automatically authenticating all users as superusers in the visual editor.
  This is a breaking change from 1.5 where both of these environment variables were true by default, bypassing all authentication.
  For temporary backwards compatibility, you can revert to the earlier unauthenticated behavior by setting both variables to true. However, a future release will set LANGFLOW_AUTO_LOGIN=False and remove LANGFLOW_SKIP_AUTH_AUTO_LOGIN. At that point, Langflow will strictly enforce API key authentication for API requests, and you can manually disable authentication for some features, like the visual editor, by setting LANGFLOW_AUTO_LOGIN=True.
  For more information, see the documentation for LANGFLOW_AUTO_LOGIN.

New features and enhancements​

• OpenAI Responses API compatibility
  Langflow now includes an endpoint that is compatible with the OpenAI Responses API at POST /api/v1/responses. This allows you to use existing OpenAI client libraries with minimal code changes by replacing the model name with your flow_id. The endpoint supports streaming responses, conversation continuity, tool call results, and global variable passing through headers. For more information, see OpenAI Responses API.
• Advanced document parsing with built-in Docling support
  The Read File component supports advanced parsing with the Docling library.
  To make it easier to use the Docling components and the Read File component's new advanced parsing feature, the Docling dependency is now included with Langflow for all operating systems except macOS Intel (x86_64).
  For more information, see Advanced parsing.
• Reorganized component menus and visual editor controls
  • The workspace sidebar is divided into separate sections for Search, Core components, MCP servers, Bundles, and Add Note.
  • Lock/unlock controls moved to flow details in Projects.
  • Zoom and help controls moved to the lower-right corner of the workspace.
  • Vector store components moved to provider-specific Bundles
  • Serper Google Search API component moved to the Serper bundle
• Increased the default maximum file upload size from 100 MB to 1024 MB.
• New integrations and bundles:
  • Traceloop
  • Cohere Rerank
  • S3 Bucket Uploader
  • Composio single-service components

Deprecations​

• The Local DB component is now in legacy status. Replace this component with the Chroma DB component.

Highlights of this release include the following changes. For all changes, see the Changelog.

New features and enhancements​

• Langflow API requests can require authentication
  To enhance security and ensure proper authentication for automatic login features, Langflow API endpoints now require authentication with a Langflow API key, even when LANGFLOW_AUTO_LOGIN=True. This change will be enforced in a future release. For temporary backwards compatibility, this release adds the LANGFLOW_SKIP_AUTH_AUTO_LOGIN environment variable. The default value is true, which disables API authentication enforcement. To enforce API authentication, set LANGFLOW_SKIP_AUTH_AUTO_LOGIN=False. For more information, see the documentation for LANGFLOW_AUTO_LOGIN.
• Centralized Language Model and Embedding Model components
  The Language Model component and Embedding Model component are now core components for your LLM and embeddings flows. They support multiple models and model providers, and allow you to experiment with different models without swapping out single-provider components. Find them in the visual editor in the Models category.
  The single-provider components moved to the Bundles section. You can use them to replace the Language Model and Embedding Model core components, or connect them to the Agent component with the Connect other models provider option.
• MCP server one-click installation
  On your Langflow project's MCP server page, click Auto install to install your Langflow MCP server to MCP clients with just one click. The option to install with a JSON configuration file is available for macOS, Windows, and WSL. For more information, see Use Langflow as an MCP server.
• MCP server management
  You can now add, remove, and edit your MCP servers in the MCP Tools components and through your Langflow Settings page. For more information, see Use Langflow as an MCP client.
• Input schema replaces temporary overrides
  The Input schema pane replaces the need to manage tweak values in the API access pane. When you enable a parameter in the Input schema pane, the parameter is automatically added to your flow's code snippets, providing ready-to-use templates for making requests in your preferred programming language.
• Tools components are redistributed
  All components in the Tools category were moved to other component categories, such as Helpers and Bundles, or marked as legacy.
  The MCP Tools component is now under the Agents category.
  Tools that performed the same function were combined into single components that support multiple providers, such as the Web Search component and the News Search component.
• Stability improvements
  General stability improvements and bug fixes for enhanced reliability. See an issue? Raise it on GitHub.
• New integrations and bundles
  • Cleanlab bundle

Highlights of this release include the following changes. For all changes, see the Changelog.

New features and enhancements​

• Enhanced file and flow management system with improved bulk capabilities.
• Added the BigQuery component
• Added the Twelve Labs bundle
• Added the NVIDIA System-Assist component

Deprecations​

• Deprecated the Combine Text component.

Highlights of this release include the following changes. For all changes, see the Changelog.

New features and enhancements​

• Added an enhanced Breaking Changes feature to help update components without breaking flows after updating Langflow.

Highlights of this release include the following changes. For all changes, see the Changelog.

New features and enhancements​

• Introduced MCP server functionality to serve Langflow tools to MCP-compatible clients.
• Renamed Folders to Projects in the visual editor.
• The /folders endpoints now redirect to /projects.

Deprecations​

• Deprecated the Gmail, Google Drive, and Google Search components. For alternatives, see the Google bundle.

See the Changelog.