Talk2View Integrations
Experience Talk2View embedded directly into the tools you already use. Browse our integrations and get started in minutes.
3D Slicer Medical Imaging Platform
Bring the Talk2View AI assistant into 3D Slicer, the free, open-source platform for medical image analysis and scientific visualization. Talk to your data, run AI segmentation, and orchestrate complex imaging workflows in plain language.
Natural-language control
Describe what you want in plain language and the Talk2View agent orchestrates 3D Slicer for you — no menu-hunting.
AI segmentation
State-of-the-art segmentation with TotalSegmentator, nnInteractive, MONAI, and an adaptive brush — driven conversationally.
IDC & DICOM
Search and load studies from the NCI Imaging Data Commons and work with your own DICOM data.
Voice dictation
Hands-free, on-device speech-to-text (Whisper) so you can talk to your data while you work.
Multilingual
Interact in your own language — the assistant understands and responds in many languages.
Extensible agent
Add and manage agent skills, connectors, and tools to teach the assistant new workflows.
Download
Disclaimer
Not a medical device. For educational and research use only. See Terms of Use.
Compatibility: This extension requires 3D Slicer version 5.10.0. Need a different version? Contact our team.
Installation
- 1Download the Talk2View extension for your operating system.
- 2In 3D Slicer, open the Extension Manager (documentation).
- 3Click Install from file… and select the downloaded file.
- 4Restart 3D Slicer.
- 5In the Talk2View panel, sign in with your email and password.
- 6Select a sample data set and click Download, or load your own data using the Data module.
- 7Start talking with Aimelia, the Talk2View assistant, to explore your data.
Build and debug your own AI imaging agent
The extension is a full agent platform. Inspect every step of a run, extend the assistant with your own skills and tools, plug in the model you prefer, and call best-in-class segmentation — all inside 3D Slicer.
LLM Debug Inspector
Step through every agent run and inspect the raw messages, tool definitions, tool calls, responses, timing, and token usage — pause, continue, or break at any step.
Skills, connectors & tools
Teach the agent new workflows with Markdown skills and Python connectors that expose tools. Enable, disable, duplicate, import, and export them from a built-in manager.
Bring any model
Use Anthropic Claude, OpenAI, and Google Gemini, plus medical-specialized models like Google MedGemma. Available models depend on your plan.
State-of-the-art segmentation
nnInteractive, TotalSegmentator, MONAI Label, MONAI Auto3DSeg, and an adaptive brush — all callable as agent tools.
Supported models
Segmentation
More information — extending the agent
Agent skills
Skills are simple Markdown files (with YAML frontmatter) that teach the assistant how to perform a workflow. Drop new ones into your user skills folder or create them in the built-in manager, then enable, disable, duplicate, or import/export them as .zip.
Connectors & tools
Connectors are Python classes that expose tools to the agent. A tool is just a typed method whose docstring becomes its description for the model — dependencies install automatically. Built-in connectors include the Imaging Data Commons browser, MedGemma, and every segmentation engine above.
LLM Debug Inspector
Open the inspector to see a timeline of every exchange: messages sent to the model, available tools, the model’s response and tool calls, metadata, and token usage. Step, continue, or break to debug agent behavior live.
OHIF Web-based Image Viewer
Bring the Talk2View AI assistant into OHIF, the open-source, zero-footprint web viewer for medical imaging. Talk to your scans, run AI segmentation, and orchestrate radiology workflows in the browser — nothing to install.
Natural-language viewer control
Ask the assistant to set window/level, navigate slices, change layout, and apply hanging protocols — it drives the viewer for you.
Vision-language Q&A
Ask questions about the scan currently on screen; answers are powered by Google MedGemma.
AI segmentation
Automated organ and anatomical segmentation with TotalSegmentator, plus OHIF’s brush, eraser, and threshold tools.
IDC & DICOMweb
Load studies from the NIH Imaging Data Commons or connect any DICOMweb server.
Voice dictation
Multilingual speech-to-text for hands-free notes and prompts.
Markdown reporting
Turn AI analysis and your own notes into clean, formatted reports.
See it live at app.talk2view.com
app.talk2view.com is our demo example app — the Talk2View web viewer built on OHIF. Load a public study from the Imaging Data Commons or upload your own DICOM, then talk to Aimelia, the Talk2View assistant.
It’s free to start — sign up to use the assistant, and for more usage tokens you can upgrade your plan.
Add Talk2View to your own OHIF deployment
The Talk2View web viewer is a custom OHIF 3.x deployment with drop-in extensions and a mode. Link the prebuilt panels into your own OHIF build, or embed the assistant anywhere with the @talk2view/sdk.
Drop-in OHIF extensions & mode
Prebuilt chat, transcription, segmentation, and Imaging Data Commons panels plus a Talk2View mode you can link into an OHIF 3.x deployment.
Built on @talk2view/sdk
Wrap your panel with T2VProvider and use the useT2VAuth, useT2VChat, and useT2VTools hooks.
Your own tools
Register typed client tools so the assistant can read viewer state and drive your UI in natural language.
Bring any model
Models are managed by the Talk2View platform — Anthropic Claude, OpenAI, Google Gemini, and Google MedGemma.
Supported models
More information — integrating with OHIF
- 1Install
@talk2view/sdkand wrap your chat panel withT2VProvider, passing your partner key. - 2Use the useT2VAuth, useT2VChat, and useT2VTools hooks to handle sign-in, messaging, and tool registration.
- 3Register typed client tools that read viewer state and control Cornerstone — window/level, navigation, layout, hanging protocols, and segmentation.
- 4Or skip the wiring and link the prebuilt OHIF extensions (chat, transcription, segmentation, Imaging Data Commons) and the Talk2View mode into your deployment.
The Joy Matrix
A productivity tool that extends the Eisenhower Matrix for teams, scoring tasks by pleasure, talent, and capacity to move from A to B without burning anyone out. It’s free to use and powered by Talk2View.
Open source, in the public domain
The Joy Matrix is released into the public domain under CC0 1.0 Universal. Use it, fork it, sell it, or ship it as your own — no attribution required, no permission needed, no strings.
Features
Urgency × importance matrix
Four quadrants — Do, Schedule, Delegate, Eliminate — extending the classic Eisenhower method.
Multidimensional scoring
Tasks scored by pleasure, talent, and capacity — not just urgency and importance.
Smart auto-assignment
Work is routed to the right teammate based on talent and available capacity.
Burnout prevention
Capacity ceilings and load balancing keep the plan sustainable — from A to B without burning anyone out.
Team insights
See each person’s load, joy index, talent fit, and burnout risk at a glance.
Scheduling
Turn due dates and availability windows into a workable schedule.
Five focused views
Matrix, Team, Tasks, Schedule, and Insights.
Talk to your tasks
Ask the built-in Talk2View assistant things like “Who is at burnout risk?”
Themes & customization
Light and dark modes with customizable presets.
Import & export
Portable JSON in and out, plus one-click demo data.
Local-first & private
Runs entirely in your browser with no signup — your data stays with you.
Using Talk2View in The Joy Matrix
The Joy Matrix is free to use with or without Talk2View. The matrix, scheduling, and insights all run on their own with no account required.
The conversational assistant is powered by Talk2View. Sign up to talk to your tasks inside The Joy Matrix — it’s free to start, and for more usage tokens you can upgrade your plan.
A working demo of the Talk2View SDK
The Joy Matrix doubles as a reference implementation for adding Talk2View to an existing or new app with the @talk2view/sdk. Register your app’s actions as tools, drop in the chat widget, and the assistant can drive your UI in natural language.
More information — how the integration works
- 1Install
@talk2view/sdkand renderTalk2ViewandChatWidgetfrom@talk2view/sdk/ui, passing your partner key. - 2Define client tools that let the assistant read and mutate your app state, with helpers for fuzzy name matching.
- 3Teach the assistant your domain with a system prompt and skills written in Markdown.
The integration code
- JoyMatrixChat.jsxRenders Talk2View and the ChatWidget from @talk2view/sdk/ui with your partner key.
- tools.jsClient tools that let the assistant read and act on your app’s state.
- lookup.jsFuzzy name resolution used by the tools.
- system-prompt.mdThe assistant’s system prompt.
- skills/joy-matrix.mdA reusable skill that teaches the assistant the domain.
LibreOffice Writer
Bring the Talk2View AI assistant into LibreOffice Writer, the free, open-source word processor and alternative to Microsoft Word. Ask it to write, edit, format, restructure, and review your document in plain language — powered by leading AI models including Anthropic Claude, Google Gemini, and OpenAI GPT.
Natural-language editing
Ask the assistant to write, rewrite, and restructure your document — it makes the edits for you, right in Writer.
Sidebar chat panel
A Talk2View chat deck lives in the LibreOffice Writer sidebar, always available alongside your document.
Formatting & styles
Apply text and paragraph formatting and manage document styles through natural-language requests.
Tables & images
Insert and edit tables and insert images without leaving the conversation.
Structure & search
Manage headings, page breaks, sections, and the document outline, plus find & replace across the document.
Comments & review
Add, edit, and resolve comments, and run a pre-send review before the document goes out.
Get the extension
Talk2View for Writer ships as a single universal .oxt package — one file that works on Linux, macOS, and Windows. Install it from Tools → Extension Manager → Add… or with unopkg add --force Talk2ViewWriter.oxt, then restart Writer and the Talk2View deck appears in the sidebar.
Requires LibreOffice 7.0 or later and a Talk2View account. It’s free to start — sign up to use the assistant, and for more usage tokens you can upgrade your plan.
Checking for the latest release…
Open source on GitHub
Talk2View-Writer is an open-source LibreOffice extension: a React + @talk2view/sdk chat panel wired to a catalog of Python tools that drive Writer through the UNO API. Read the code, build it yourself, or contribute.
Open source (MPL 2.0)
Released under the Mozilla Public License 2.0 — the same license LibreOffice itself ships under. Read the code, fork it, and contribute.
Built on @talk2view/sdk
The chat UI is a React bundle using @talk2view/sdk/ui; auth, chat, and tool registration all run client-side through the SDK.
Python UNO tool catalog
20 tools across reading, writing, formatting, search, structure, and commenting, implemented against the LibreOffice UNO API.
Cloud engine backend
Talks to the Talk2View cloud engine with a partner key and user login — models are managed by the platform.
Supported models
More information — installing & building
- 1Download the latest
Talk2ViewWriter.oxtfrom the GitHub Releases page. - 2Install it from Tools → Extension Manager → Add… in LibreOffice, or run
unopkg add --force Talk2ViewWriter.oxtfrom the command line. - 3Restart Writer — the Talk2View deck appears in the right sidebar.
- 4Sign in with your Talk2View account and start asking the assistant to work on your document.
T2Board
AI chat, right in your keyboard. T2Board adds a custom keyboard to your device so you can prompt leading AI models and drop the answer into any app — powered by your Talk2View account.
Type anywhere with the globe key
Switch to T2Board with the globe key in any app, type a prompt, and get an AI response — without leaving the app you're in.
Your choice of models
Pick your provider and model — OpenAI, Anthropic, Google, and OpenRouter — grouped by provider for quick switching.
Streaming, formatted replies
Answers stream in with rich formatting — bold, italic, headings, lists, and code blocks all render inline.
Built for messaging
Drop AI responses straight into the conversation in WhatsApp, iMessage, and any other text field.
Select, copy & look up
Message text stays fully selectable, with copy and look-up so you can reuse any part of a reply.
One Talk2View account
Sign in with your Talk2View account — the same one that powers every other Talk2View integration.
Get T2Board
T2Board is a free download. Install it, enable the keyboard, and sign in with your Talk2View account.
New to Talk2View? Create a free account to sign in. See the privacy policy and terms of use.