Changelog

Today, we’re introducing Langdock for Excel, available on the Microsoft Marketplace. Langdock now lives directly inside Excel, so everything you need is right there in your spreadsheet.

Why we built this

Spreadsheets are where much of the work happens, but switching between Excel and a separate tool breaks the flow. Copying data back and forth, losing context, and constantly changing windows don’t scale when you’re trying to get real work done.

Langdock for Excel brings Langdock directly into Excel. You can read your workbook, write into it, and create new files with AI. Your existing Langdock integrations, agents, and skills are available without leaving your spreadsheet.

What you can do

• Ask questions about your workbook: get explanations of complex sheets, spot anomalies, or understand what a formula is doing without manually digging through cells.
• Create formulas, tables, charts, and pivot tables: describe what you need in plain language and let Langdock build it directly in your sheet.
• Clean and structure messy data: standardize names, dates, categories, or currencies before analysis.
• Summarize spreadsheet findings: turn data into clear text for reports, emails, presentations, or meetings.
• Save your work to the library: upload finished spreadsheets directly to your Langdock library so they’re always accessible when you or your team need them.
• Use Langdock features directly in Excel: knowledge bases, integrations, skills, and workflows are all available inside the plugin.

Availability

Workspace admins can enable Langdock for Excel at the bottom of the workspace integration settings. Once enabled, users can install it from the Microsoft Marketplace and start using Langdock directly in Excel.

You can check out our documentation to get started.

Anthropic’s Claude Sonnet 5 is now available on Langdock! 🚀

Today, we are adding Claude Sonnet 5 to Langdock. It is Anthropic’s most agentic Sonnet model yet. It can make plans, use tools, and run autonomously at a level that previously required larger and more expensive models. Its performance is close to Opus 4.8, but at a lower price.

You can read more about the launch in Anthropic’s official announcement.

The model is hosted in the EU and available to all workspaces now.

To get you started quickly, we have automatically enabled Sonnet 5 in all workspaces where a previous Sonnet version was activated. Admins can manage model access in workspace settings.

You can now test your agents before shipping them, directly inside the agent editor.

Agent Evals lets you define test sets with real prompts, set expectations for what a good answer looks like, and run them against your agent on demand. No more manual spot-checking after every update.

What's new

Test sets & test cases

Create one or more test sets per agent. Each test set holds a collection of test cases - a prompt paired with the criteria the agent's response must meet. You can add cases manually, edit them inline in a table view, or import them in bulk via CSV (a sample template is available to download).

Three ways to grade a response

• AI judge: Provide a reference answer or a rubric, and a model evaluates whether the agent's response passes. The judge returns pass, unsure, or fail.
• Tool use check: Specify which tools the agent should call. Pick from the agent's connected tools and actions using a built-in tool picker.
• Keyword check: Define words or phrases the response must mention or must not mention.

Choose your judge model
Each test set lets you pick which model grades the responses, independent of the model the agent itself uses.

Run and watch results live
Kick off a run from the Tests tab. Results stream in as cases complete. A cross-set timeline gives you a history of all past runs. Click any case in the results table to open a full detail view: the conversation, tool inputs and outputs, and the expected vs. actual answer side by side.

Run individual cases
You don't have to run an entire test set every time. Select one or more cases and run only those - useful when iterating on a specific prompt. Unselected cases are skipped and stay in the set.

Download results as CSV
Export any run's results as a CSV for analysis or sharing with stakeholders.

Safe tool execution

When running tests, tool calls that could have side effects (like writing data or sending messages) are executed in a safe mode by default, so you can test freely without worrying about unintended actions.

Note: Test runs count against your personal usage limits.

Where to find it

Open any agent in the editor and go to the Evals tab. The tab was previously visible but disabled, it's now fully available.

We'd love to hear how this fits into your workflow of building agents!

Today, we are introducing Work in Folders, which gives your chat a working directory. Choose a folder, and Langdock can use the files inside it while creating, editing, and saving new work back to the same place.

Join our webinar

We will host a webinar on July 1st to walk you through the different ways to use files in Langdock. Get a first hand introduction to how files, knowledge bases, folders, templates, and the new document editor work in day-to-day use. Sign up here.

Why we built this

With the Library, we made it much easier to find and manage files in Langdock. Work in Folders takes the next step by bringing those files directly into the way you work with AI.

Instead of attaching files one by one or starting from an empty chat, you can select a folder and give Langdock the right context from the start. The same folder also becomes the place where new or edited files are saved, so folders are not just for storing work, but for actively working with it.

What's new

• Work from a folder in chat: Select "Work in a folder" before starting a task. Langdock can use the files inside to answer questions, summarize, compare, or create new work, while generated and edited files stay connected to the same folder.
• Collaborative folders: Folders can be shared with colleagues, groups, or your whole workspace, making them useful for teamwork, recurring tasks, and shared file collections.
• Use folders in Projects: Work in Folders is also available in Projects. Add a folder to give the Project lasting context, and keep new or edited outputs in the same place.
• Use folders in Agents: Folders are also available in Agents as read-only folders for now.

For example, to prepare for a customer meeting, put proposal notes, usage numbers, financials, call notes, and a presentation into one folder, then ask Langdock to turn it into a clear agenda or first draft. The context is already there, and the output lands right where you need it.

Folders and Knowledge Bases

This release also makes the difference between folders and knowledge bases clearer. The folders you already had in Langdock now appear as knowledge bases, while the new folders are built for active file work.

• Knowledge bases are read-only collections for semantic search, and reusable company knowledge across chats and agents. Existing folders keep their content and settings, but now appear as Knowledge Bases.
• Folders are for active work. Store files, share them, and work with them in chat while Langdock reads, edits, and creates files alongside you.

You can find more information in our documentation.

Head to the Library, create a folder, add your files, and select Work in a folder in chat to get started!

Scheduled Tasks lets you set up a prompt that runs automatically on a schedule you define. Combine it with an agent and your integrations, and Langdock handles the rest, whether that’s a daily briefing, a weekly summary, or any other task you'd otherwise do manually on a regular basis.

What's new

Recurring prompts on your schedule
Create a scheduled task by writing a prompt, choosing an agent, and deciding when it should run. Schedules can be set to run daily, weekly, or at a custom cadence. Each run automatically starts at the time you set, in your local timezone.

Powered by your agents and integrations
Scheduled tasks can run through any agent you've already configured in Langdock. Your daily Slack digest, weekly CRM summary, or automated report can all pull live data from the integrations your agent already has access to.

A dedicated place for all your scheduled tasks
A new Scheduled section in the sidebar gives you a central place to see, manage, and monitor all your tasks. You can see the active tasks directly in the sidebar, so you always know what's running in the background.

See previous run history
You can review past runs, see what the agent produced, and adjust the configuration if needed.

Pause and resume anytime
You can easily pause tasks with one click. Paused tasks are clearly marked and won't continue until you restart them.

How to get started

• Go to “Scheduled” in the sidebar
• Click “New task”
• Write your prompt: describe exactly what you want the task to do
• Choose an agent (optional)
• Set your schedule: pick the frequency it should run
• Save: your task will run automatically from that point on

Example schedules to get you started

Here are a few ideas to help you schedule your first task:

• Daily standup prep: Every weekday at 8:30 AM, pull your open Jira tickets, recent GitHub PRs, and Slack mentions into a concise "here's what I'm working on today" summary, ready before your standup starts.
• Weekly competitive pulse: Every Monday morning, scan a set of competitor blogs, product pages, or news sources and deliver a short briefing on what changed: new features, pricing updates, announcements.
• End-of-week team digest: Every Friday at 4 PM, summarize the week's closed tickets, merged PRs, and key decisions from Notion or Confluence into a digest you can paste straight into your Teams channel.
• Daily CRM hygiene check: Each morning, flag any deals in your CRM that haven't had activity in over a week, and draft a short follow-up message for each one, ready for you to review and send.
• Monday metrics briefing: Pull last week's key numbers from your analytics tool or data warehouse and format them into a short report with highlights and anything that looks off.
• Async meeting prep: The night before a recurring meeting, gather the latest updates from relevant Confluence pages, Google Docs, or Notion pages, and draft a pre-read.

We are introducing a context window indicator directly in the chat input bar, so you can see exactly how much of a model's context is being used, and what is using it.

Why we built this

Every AI conversation has a capacity limit (called the context window), a fixed amount of information the model can work with at any one time. As conversations grow, older messages are automatically summarized to free up space. Until now, there was no easy way to see how full the context window was or what was actually filling it up. This feature gives you visibility into what's going on behind the scenes and how much of the context window each part of the conversation uses.

What you can see

The indicator sits in the chat input bar and shows your current context window usage, for example, 13.6k / 200.0k (7%). You can click it to open a detailed breakdown.

The breakdown groups context usage into the following categories:

• System tools: tools available to the model in this chat
• System prompt: core instructions for the model
• Messages: the visible conversation between you and the model
• Tools: the built-in or integration actions performed in the chat
• Attachments/images: files and images you have added to the chat
• Skills: active skill descriptions available to the model
• Memory: saved memories included in the chat
• Autocompact summary: a summary of earlier messages
• Free space: how much context remains available

Each category shows a token count (the unit AI models use to measure text), a percentage, and a position in a stacked visual bar.

You can find more information about this in our documentation.

We would love to hear how this works for you 🙌

Anthropic's Claude Opus 4.8 is now available on Langdock! 🚀

Today, we are adding Claude Opus 4.8 to Langdock. It builds on Opus 4.7 with improvements across benchmarks and is Anthropic's strongest model for long-running agentic tasks. It is designed to be a more effective collaborator: it carries context better, catches its own mistakes, and pushes back when a plan doesn't sound right.

You can read more about the launch in Anthropic's official announcement.

The model is hosted in the EU and available to all workspaces now.

To get you started quickly, we have automatically enabled Opus 4.8 in all workspaces where another Opus version was previously activated. Admins can manage model access in workspace settings.

Google's Gemini 3.5 Flash is now available on Langdock! 🚀

Today, we are adding Gemini 3.5 Flash to Langdock. It is Google's fastest and most capable Flash model yet, built for agentic tasks, coding, and multimodal work.

You can read more about the launch in Google's official announcement.

The model is hosted in the EU and available to all workspaces now. You can also enable "Thinking" directly in the model dropdown for complex tasks that benefit from deeper reasoning.

To get you started quickly, we have automatically enabled Gemini 3.5 Flash in all workspaces where a previous Gemini model was activated. Admins can manage model access in workspace settings.

We've launched a new document editor in Langdock: a redesigned way to write, edit, format, and collaborate with AI on documents, right next to your chat.

From Canvas to a real document editor

We've built a completely new document editor, integrated directly into the chat, that replaces the previous Canvas functionality. The core experience stays the same. The main shift is moving from a canvas toward creating real documents.

With the recent launches of Skills, Templates, and Library, we began introducing a new way of working with files in Langdock. The shift from "Canvas" to "Creating Documents" is also the foundation for further enhancements in the coming weeks, focused on working with files and folders more naturally in Langdock.

What's new

Create documents naturally

Click the "Create document" button (previously "Canvas") or just describe what you want in chat (e.g., "write me a guide on X", "draft a proposal"). Langdock will spin up a new document and start drafting right alongside you.

Refreshed design and richer formatting

A clean interface with all the formatting options you'd expect: headings, lists, quotes, code blocks, bold, italic, underline, strikethrough, and inline links. Switch to full-screen view when you want to focus on the document, or keep the chat alongside as you work.

Save to SharePoint, OneDrive, and Google Drive

Send finished documents straight from the editor to your connected tools, so they land where the rest of your work already lives. No more download-then-upload.

Hand in hand with the Library

Documents you create in Langdock now live in your Library and can be opened and edited there, not just in the chat where they were created. Head to your Library, open any document, and pick up where you left off.

Version history

Every update, whether made by you or the AI, is saved as a new version. Browse the full history, preview older versions, and restore any of them with a click.

Note: editing currently supports Langdock documents only, not Word or PDF files.

What happened to the old Canvas?

The old Canvas is now fully replaced by the new document editor, which covers what Canvas did and adds quite a bit on top. Any existing Canvas documents remain accessible in the original chats where they were created. Good to know: for a smoother transition, Langdock still recognizes "Canvas" in prompts and opens the new document editor.

Start a new chat, click on "Create document", and give it a try.

We can't wait to hear what you think! 🙌

Auto is our new model routing option that automatically selects a suitable model for your task.

Instead of manually choosing between different models for every prompt, Auto analyzes the first message of a conversation to understand the request and estimate its complexity. Based on that initial assessment, it selects a model for the rest of the conversation.

The goal is not to replace manual model selection entirely. You can still switch to other models at any point in a conversation. Auto is designed primarily for everyday work: quick questions, drafting, summarizing, rewriting, analysis, and other common tasks where you should not have to think about which model is suitable.

By taking away this small but frequent model-choice decision, Auto helps you work faster and make better use of your usage limits by avoiding costly models for simpler tasks. We will continue to improve the routing logic over time, expand the available model choices, and use feedback from users to make the experience more accurate and flexible.

Usage transparency in the model selector

We've added a usage bar directly in the model selector, so you can see at a glance how much usage roughly each model consumes. This makes it easier to understand your limits, make informed model choices, and get the most out of Langdock.

OpenAI's GPT‑5.5 is now available on Langdock! 🚀

Today, we are adding GPT‑5.5 to Langdock. It's OpenAI's smartest model yet and brings improved understanding of intent and the ability to plan, research, and work through complex tasks on its own.

You can read more about the launch in OpenAI's official announcement.

The model is hosted in the EU and available to all workspaces now.

To get you started quickly, we have automatically enabled GPT‑5.5 in all workspaces where GPT‑5.4 was previously activated. Admins can manage model access in workspace settings.

Agents now support versioning and richer analytics, giving builders full control over when changes go live and deeper insight into how their agents are used.

Agent Versioning

Agents now have versions, so you can iterate with confidence and never lose a working configuration.

When you edit an agent, your changes are saved as a private draft that only you can see. Nothing goes live until you're ready. When you click "Update" in the top right of the screen, your draft is published as a new version, instantly available to everyone with access to the agent.

What's new:

• Drafts autosave as you edit: experiment freely without affecting anyone using the agent
• Publish when you're ready: a single "Update" click rolls out your changes to all users
• Every update is saved: a full version history is kept, so you can browse past versions and restore any of them at any time
• Change summary on publish: see the difference of what exactly has changed (e.g., name updated, instructions updated, actions added), and optionally add a short description to communicate the changes

Agent Analytics

Agent owners and editors now get a much richer, more actionable view into how their agents are performing.

The analytics experience has been moved to a top-level tab inside the agent, making it significantly more prominent and easier to access. The new interface brings deeper charts and a more powerful feedback review flow, so that you can continuously improve your agents.

What's new:

• Richer usage charts: get more detailed metrics on conversations, active users, and engagement over time
• Improved feedback interface: browse and filter user feedback, switch between conversations, and preview chats without leaving the analytics page
• More prominent placement: analytics accessible directly from the top navigation bar within the agent

Together, these improvements give agent builders the visibility they need to understand what's working, catch issues early, and continuously improve their agents based on real usage.

Who is this for?

Both features are available to agent owners and editors. End users of agents are unaffected; they always see the latest published version, with no visibility into drafts.

We added two new models to Langdock today: Claude Opus 4.7 and GPT Image 2. 🚀

Claude Opus 4.7 is Anthropic's latest flagship model and its most capable available model to date. It brings stronger performance across coding, vision, and complex multi-step tasks. Whether you are tackling large projects, long-running workflows, or sophisticated agents, Opus 4.7 handles them with greater precision and consistency than its predecessor. You can read more in Anthropic's official announcement.

GPT Image 2 is OpenAI's next-generation image model and a major step forward for visual generation. It delivers 2K resolution across multiple aspect ratios, significantly improved text rendering in images, and real-world knowledge updated through December 2025. You can read more in OpenAI's official announcement.

To get you started quickly, we have automatically enabled both models in all workspaces where Opus 4.6 / GPT Image 1.5 was previously activated. Admins can manage model access in workspace settings.

We're introducing our Library – one central place for everything file-related in Langdock. Create new files, manage folders, browse every file you've generated or uploaded, and use file templates to keep output consistent.

Join our webinar

We will host a webinar on April 14th to walk you through our new Library. Get a firsthand introduction on how the Library works and you can use it in your day to day work. Sign up here!

Why we built this

As more teams use Langdock to create documents, presentations, and work with uploaded files, finding them again has become a common pain point. Files were tied to the conversations they were created in, and there was no single place to browse, manage, or reuse them. The Library changes that by bringing all file-related functionality into one view – so you spend less time searching.

What's in the Library

The Library includes three sections:

• Create New File – Start a new document, spreadsheet, or presentation directly from the Library. Files can be created blank or from your team's templates.
• Folders – What was previously called "Knowledge Folders" is now just called "Folders" and lives inside the Library. Same functionality, new home. Create folders, upload files, share with colleagues, and attach them to agents or reference them in chat via @mentions. All your existing folders and settings carry over automatically.
• Recent Files – Every file you've created, uploaded, or imported across all your Langdock conversations appears here. Preview files, download them, jump back to the original chat, or start a new chat with any file attached. They can now easily be moved into folders. Good to know: these files follow your workspace's chat data retention settings. When a chat is deleted, the associated files are removed as well.

File Templates

File templates let you save existing documents or slide decks as reusable starting points. Upload your company's master PowerPoint presentation, offer template, or any document that follows a consistent format – and anyone you share it with can generate new files in the same structure, layout, and style.

Upload a template from your computer or import one from a connected integration, such as Google Drive or SharePoint (supported files are .pptx, .pdf, and .docx). Langdock analyzes the file to make it ready to be used in Chat. Templates can be shared with specific users, groups, or workspace-wide.

Head to the Library in your sidebar to get started, and sign up for our Library webinar here!

We’re introducing the Get Started Guide – a personalized onboarding experience that helps every user interactively discover the Langdock platform.

When teams adopt Langdock, users often start with chat and gradually discover the rest of the platform on their own. The Get Started Guide helps here by giving every user a checklist of tasks across the entire Langdock platform, with points and progress tracking to keep things engaging.

How it works

When you open Langdock, you will find a new Get Started page with tasks organized by product area: Chat, Prompt Library, Projects, Agents, Knowledge, Integrations, and Workflows. Each task walks you through a key feature – from sending your first message to creating an agent with subagents.

Complete tasks to earn points and climb your workspace leaderboard. The guide adapts to each user so they only see the features most relevant to them and the ones they have access to. Admins can also see workspace setup tasks.

Spotlight tours

Many tasks include interactive spotlight tours that guide you step-by-step through the interface, so you always know exactly where to go and what to do next.

Leaderboard

A workspace leaderboard lets you see how your team is progressing. Admins can choose to anonymize leaderboard entries if needed.

Custom tasks

Admins can add custom onboarding tasks from the workspaces settings to guide teams through company-specific steps – for example, "Read our AI usage policy".

Who sees it

The Get Started Guide is automatically enabled for users who joined recently. If you have been using Langdock for a while, you can access the guide at any time by heading to your account settings – a useful way to revisit features you may not have explored yet.

Let us know how the guide works for your team 🙌