Tao AI Studio is an agent-controlled desktop work environment. Use local or cloud models to plan and ship software, research the web, create documents, automate Windows workflows, and organise personal work — with persistent memory, parallel agents, MyLife, and an extensive built-in toolset.
Windows 10/11 · .NET 9 Runtime included · Free
Tao AI Studio is a single self-contained Windows executable. No package manager, no runtime to install, no configuration files to hand-edit before you can start.
From code editing and AI chat to research, Office document generation, MyLife, desktop automation, and autonomous agents — all in a single desktop application.
Persistent episodic memory with vector search. Memories are organised into Rooms, Halls, Chains, and Tunnels — surfacing the right context every time.
Spawn independent background worker agents that run up to 40-turn agentic loops concurrently. Monitor live progress without blocking the main chat.
Edit spreadsheets with natural language. Type "Create a budget tracker" and watch the AI generate formulas, data, and formatting automatically.
Automatic self-review loop that keeps agents accountable. The agent evaluates its own work against project tracker goals and continues until satisfied.
200+ available tools organised into categories that auto-expire after 10 seconds of inactivity, keeping every LLM call lean and focused.
Switch between Ollama, OpenRouter, vLLM, LocalAI, HuggingFace, and Llama.cpp with a model prefix. Unified message format across all providers.
Full-featured AvalonEdit integration with syntax highlighting, multi-tab support, and automatic file type detection across 15+ languages.
Generate class, sequence, component, and namespace summary diagrams. Pre-generated diagrams can be injected into the system prompt for architectural awareness.
Vision model support for image analysis plus text-to-image generation. Attach images, videos, audio, and documents as conversation context.
Transform any image into a 3D model using HuggingFace Trellis or Stability AI's Stable Fast 3D. Output GLB files and view them in the built-in Three.js 3D viewer tab.
Model Context Protocol integration — connect any MCP server and its tools are automatically discovered and merged with the built-in tool library.
Enable reasoning mode on supported models with a configurable 1K–32K token budget. Works with OpenRouter thinking models, vLLM, and llama.cpp /think commands.
A structured planning interface lets the agent outline its approach before executing. Working memory gives agents a persistent scratch pad across turns.
Integrated Brave Search API gives agents real-time internet access with 2,000 free calls/month. Paired with a web-fetch tool for full page content extraction.
Full IMAP/SMTP email client built in with a clean tap-to-navigate layout — click a message to open the full email view, ← Back to return to the list. Works with any provider (Gmail, Hostinger, iCloud). Agent tools can scan, organise, reply to, extract contacts and events from, and save attachments from email. Peek mode browses without affecting server state.
Per-session state tracker injected into every agent turn — current goal, active tasks, completed tasks, recent errors, and agent-written findings. Auto-persisted to chat history.
Full project snapshot taken before every Agent Chat run. Browse and restore any of the last 10 snapshots from Workspace → Checkpoints if an autonomous run goes wrong.
Agents auto-generate reusable procedure files from successful runs. Top-3 matching skills are injected before each run; agents can query mid-run with check_skills(). Stale skills pruned automatically.
Agents can schedule themselves to run again via setup_cron_job — daily, weekly, or on an interval. Jobs appear in Workspace → Cron Jobs and fire as parallel agents automatically.
7 CDP-based tools give agents full control of the in-app WebPage tab — navigate, click, type, scroll, run JavaScript, and extract text. Full write → open → interact → screenshot → verify loop.
Condition the AI's response style using one of 16 MBTI personality archetypes — from INTP Critic to ENFJ Protagonist. Select from the Advanced toolbar or set a default in Settings.
A row of coloured icon buttons in the header bar gives one-click access to your most-used workspaces. Right-click to remove or reorder; the + button adds from 29 base workspaces. Enterprise Policy and Org Dashboard appear when configured. New installations receive a default set, and all later changes are saved instantly.
Markdown note editor with folders, tags, pinning, and dual-scope storage (global + per-project). Two-phase loading for instant lists, debounced auto-save, and 6 agent tools for autonomous note management.
Keep private personal details and lasting global or project todos in one place. Send individual details to the Windows-encrypted Credential Vault; agents can manage todos and see detail labels, never their values.
Personal address book with coloured avatar initials, multiple emails/phones/addresses, tags, and company fields. Stored per-project or globally. 4 agent tools for AI-driven contact management.
Month-view calendar with mini-grid, coloured event dots, upcoming events list, and full event form (all-day, recurrence, attendees, reminders). 6 agent tools including in-range and keyword search.
File store for receipts, invoices, contracts, reports, photos, and more. Category sidebar, drag-and-drop upload, metadata fields (vendor, amount, date, tags). 5 agent tools including text-content reading.
This is an agent-controlled work environment, not a chatbot with a few extras. These are real workflows you can run from a single instruction and review as they happen.
Open an Agent Chat tab, enable Tools, and tell the agent to build you an HTML dashboard, game, or any web page. Here's what happens:
Open multiple AI Chat tabs and send messages in each. They stream simultaneously and independently. Compare model responses side-by-side, or run a long background task while having a quick conversation in another tab.
Ask the agent to analyse CSV files and produce a formatted report. It opens an Excel tab, sends office_command calls to build sheets, write formulas, and apply formatting — then screenshots the result to verify the layout before finishing.
Ask the agent to search the web, read source pages, download material, and turn the findings into a brief, a plan, a document, or a change in your project. It can inspect the rendered result in its in-app browser before handing it back.
Let an agent work across approved desktop tools: manage files and folders, open and focus applications, inspect system state, capture the screen, use the clipboard, and send keyboard or mouse input. Tool categories keep the scope explicit.
Tell the root agent to refactor a large codebase. It spawns three independent workers — one per module — each running a 40-turn loop simultaneously. Open the Agent Monitor and watch all three progress bars advance in parallel.
Tell the agent something important about your architecture, then close the app. Tomorrow, open a new chat and ask a related question. The Memory system retrieves the relevant context automatically — the agent already knows what you told it.
Paste in a pile of email threads and invoices. The agent reads each one, creates Calendar events for any meetings, adds senders to Contacts, files receipts in Storage with vendor and amount metadata, and writes a Notes summary pinned to your Finance folder — entirely autonomously.
Store the personal details you want close at hand, move sensitive values to the Windows-encrypted vault, and maintain global or project todos. The agent can help manage the todos without ever seeing your private values.
Ask the agent to pull your upcoming Calendar events each morning and write a daily briefing note. It calls setup_cron_job with a daily schedule — then fires automatically as a parallel agent every day without you lifting a finger.
Open any file and the right editor loads automatically. Every document type gets its own AI-integrated experience.
AvalonEdit code editor with syntax highlighting for 15+ languages
WebView2 chat with markdown rendering and streaming responses
100-turn autonomous agent with full tool access and Goal Check
Display generated and processed images; attach to chat for vision analysis
Playback MP4, AVI, MOV, MKV, WEBM directly in a tab
WAV, MP3, M4A, OGG, FLAC playback plus AI speech synthesis output
DataGrid spreadsheet with inline AI chat for natural language editing
Document tab with integrated AI chat for content generation
Slides creation and editing with AI chat assistance
View and interact with PDF documents with integrated AI chat
Rendered Mermaid diagrams — class, sequence, component, namespace
Three.js GLB viewer with pan, zoom, and rotate — outputs from Trellis/Stable Fast 3D
Embedded browser tab with back/forward/refresh navigation controls
Render and inspect SVG drawings and vector graphics inline
Tao AI Studio automatically detects file type on open and routes to the correct tab — code files open in AvalonEdit, spreadsheets launch the Spreadsheet+AI tab, images appear in the viewer, and media files play immediately.
Open any .html or .md file as raw text in the code editor. Click the filename or the eye icon in the tab bar and the file instantly opens in its dedicated preview tab — no manual conversion needed.
Switch between local servers and cloud APIs with a simple model prefix. A single unified message format keeps your code provider-agnostic. Add up to 3 custom OpenAI-compatible endpoints on top of the 6 built-in providers.
Local LLM server with native concurrent client support. The default provider — zero config needed for most setups.
Cloud gateway to hundreds of frontier models. Real-time cost and token tracking per message. Extended thinking supported with 1K–32K token budgets.
High-throughput inference server. Configurable endpoint in settings — ideal for quantised open-weight models at scale.
Self-hosted OpenAI-compatible server. Single-process — the agent semaphore serialises worker calls automatically.
Inference API integration supporting thousands of models on HF Hub. Requires an API key configured in Settings.
Direct llama.cpp server integration for maximum control over quantisation and GPU offloading on consumer hardware.
Add up to 3 custom OpenAI-compatible endpoints — Groq, DeepSeek, Fireworks, or any compatible API. Configure a display name, base URL, and API key; models appear as Name -- model-id in the dropdown.
Inspired by the Method of Loci, the Memory Palace organises knowledge spatially. Every conversation is encoded, clustered, linked, and retrievable in milliseconds — with zero extra VRAM.
Auto-clustered topic groups. Each room has a generated label from anchor keywords.
Knowledge classification: fact · procedure · event · decision
Episode links via PreviousMemoryId — walk the chain for full conversation replay.
Cross-topic bridges at 0.65 similarity — surface related ideas from other projects.
The default CurrentModel strategy reuses the active chat model for embeddings — no second model to load, no extra GPU memory consumed. Switch to a dedicated encoder for higher precision, or to keyword TF-IDF for fully offline operation.
Tao AI Studio builds and maintains a personal knowledge graph from every conversation. Use it to inspect, organise, and explicitly recall prior knowledge without silently adding inferred facts to each prompt.
Every token the AI reads costs real money. The Context Engineering Dashboard breaks the entire context window into six named components so you can see at a glance what is consuming space — and cut it before it hits your bill.
The root agent runs up to 100 turns autonomously and can spawn background workers — each running their own independent 40-turn agentic loop. The root never blocks while workers run in parallel.
launch_parallel_agentEvery Agent Chat tab has its own Working Memory — a session-scoped state block automatically injected into every system prompt so the agent always knows what it's done, what failed, and what's next. Persists across saves and restarts.
Enable the Auto checkbox and the agent runs turn after turn without waiting for your approval — up to 100 turns. Pair it with Goal Check and Checkpoints for a fully supervised hands-off run.
Each worker writes its state to .tao/agents/{name}/. The AgentStatusDialog uses a FileSystemWatcher for instant OS-level notifications — no polling, no timers.
Each provider gets a dedicated service instance per worker — no shared event handlers, no LLM response races. A single semaphore serialises calls to single-process servers like vLLM and LocalAI.
Enable Goal Check and the agent is held accountable to your active Project Tracker goal. When it stops, it must evaluate its own work — and keeps going until it's genuinely satisfied.
No tool calls detected — the agent appears to be done with its current response.
A structured review message referencing the active goal is injected as a user message — covering quality, testing, output verification, and blockers.
completed()satisfied: true — stop. satisfied: false — continue autonomously until the next review.
The cycle repeats each time the agent stops — guaranteeing it doesn't declare victory prematurely.
The agent can autonomously create, populate, verify, and screenshot Office documents — all via tool calls, with tab locking to prevent interference.
"Create a monthly budget tracker with formulas"
Current spreadsheet state + message → JSON operations from LLM
State manager applies cell edits, formulas, and formatting
DataGrid updates instantly — iterate with follow-up messages
The Tabs tool category lets agents open, command, and verify document, PDF, and web tabs. Tab locking prevents the user from interfering while the agent works.
Agents can create polished documents and presentations from scratch. Each tab includes an embedded AI chat — ask follow-up questions, request edits, or have the agent iterate on content.
Drop any 2D image and Tao AI Studio converts it to a production-ready 3D GLB model via HuggingFace Gradio Spaces — no local GPU required. The result opens instantly in the built-in Three.js viewer.
JeffreyXiang/TRELLIS — highest quality multi-view 3D reconstruction. Best for objects and products.
stabilityai/stable-fast-3d — ~30 second conversion via Stability AI's HuggingFace Space. Fast previews.
Tool categories auto-disable after 10 seconds of inactivity. Every LLM turn gets only the tools it needs — keeping context windows lean and costs down. Connect MCP servers to add more.
Core tools always available: completed(), enable_tool_category, working memory, plan mode, and agent management.
Read, write, glob, zip, hash. Directory ops, file moves, and properties. Auto-expires 10s after last use.
Brave Search API, web fetch with content extraction, URL accessibility checks, and browser bookmarks.
Launch apps, screen capture, keystrokes, mouse control, clipboard, registry, network, system monitoring.
Open and control spreadsheet, document, and presentation tabs. Screenshot verification included.
Goals, tasks, and subtasks. Git workflows are available through configured and enabled MCP servers.
Search by topic, date, or recency. Retrieve statistics and store new episodic memories mid-conversation.
Launch, poll, retrieve output from, and kill parallel worker agents. Root-agent only.
IMAP/SMTP tools to scan, read, send, reply, forward, move, delete, search, process attachments, and extract contacts or events. Any IMAP/SMTP provider — Gmail, Hostinger, iCloud. Outlook/Exchange users can add an MCP server.
Tools from any connected Model Context Protocol server are auto-discovered and added to the active set.
Convert 2D images to GLB models via Trellis or Stable Fast 3D. Open output in the built-in Three.js viewer.
Inject images into the next turn with see_image, capture screen regions, and screenshot any open tab.
Full WebPage tab control via Chrome DevTools Protocol — navigate, click, type, scroll, run JS, and extract text from any in-app browser tab.
Keystrokes, mouse control, and clipboard operations. Drive any Windows application programmatically.
Image, audio, and video processing tools. Generate images, process audio, and handle media files.
6 tools: list, read, write, append, delete, and search Markdown notes across global and project scopes.
4 tools: list, search, upsert, and delete contacts. Supports multiple emails, phones, tags, and company fields.
List, search, create, update, delete, and find conflicts across calendar events with recurrence and attendee support.
5 tools: list, search, read-text, update-metadata, and delete stored files. Vendor, amount, and document-date fields for receipts and invoices.
Manage lasting global and project todos, and list personal-detail labels with vault status. Private values remain unavailable to agents.
The Workspace Panel is a scrollable launcher of tools and utilities that operate across the whole application. Click any button to open it in a tab or dialog.
Your private details and lasting tasks. Keep global or project todos here, and move sensitive detail values into the Windows-encrypted Credential Vault.
Home overview for your personal workspaces: upcoming events, recent notes, contacts, and stored files in one quick-launch page.
Global search across Notes, Contacts, Calendar, and Storage. Available from Workspace or Ctrl+Shift+F.
Markdown note editor with folders, tags, pinning, and global/project scope. Also accessible via 6 agent tools in the Notes category.
Personal address book with avatar initials, multi-field contacts, and global/project scope. 4 agent tools for AI-driven contact management.
Month-view calendar with coloured event dots and upcoming list. 6 agent tools including keyword search and date-range queries.
File store for receipts, invoices, contracts, and documents. Category sidebar, drag-and-drop upload, and 5 agent tools with metadata support.
Live view of all running parallel background agents — status, progress bar, current tool, and turn count.
Submit issues or feature requests from inside the app. Captures recent errors automatically when logging is enabled.
Browse and restore full project snapshots taken before each Agent Chat run. Auto-pruned to 10.
Windows DPAPI-encrypted secret store for passwords and tokens, including migration support for saved email credentials.
Manage scheduled agent jobs — create, edit, enable/disable, and view run history. Agents populate this via setup_cron_job.
Add and manage IMAP/SMTP accounts used by the Email Inbox and agent email tools.
Full IMAP email client with folders, compose/reply/forward, move/delete, Peek mode, and agent-driven triage tools.
Read-only admin dashboard for active policy source, enforcement mode, spend gauges, allow/block lists, and feature status.
Side-by-side structural diff of two files with AI-assisted analysis of the differences.
Browse and select goal check review templates stored globally in TaoAIStudio/GoalChecks/.
Live application log viewer with colour-coded severity levels.
Add, configure, and manage Model Context Protocol server connections.
Browse, search, and delete episodic memory entries across all projects.
Override automatic capability detection for vision, audio, tool calling, thinking, and context limits per model.
Interactive 3D force-directed mind map of extracted knowledge — auto-built from conversations and AI-enriched for recall and inspection.
Dashboard for configured AI providers, active model state, and provider availability.
Browse and run scripts from your user TaoAIStudio/Scripts folder.
Workspace settings for role models, embeddings, providers, UI/editor preferences, files, projects, and advanced options.
Goals, tasks, and subtask manager. Agents create and update items here autonomously.
Create, browse, and activate saved system prompts for the current project.
Browse, run, and manage agent procedural skills — both auto-generated and manually authored.
Enable or disable individual agent tool categories. Usage Stats tracks token spend across all sessions.
Token usage and spend analytics by call, model, provider, session, and context source.
Optional live org-wide AI project dashboard on TaoServer, shown when configured.
IT administrators can deploy a single JSON file to lock down providers, cap spend, and disable features — without touching the application. Works with GPO, SCCM, and Intune. Users see only what policy allows.
C:\ProgramData\TaoAIStudio\ overrides any user-level fileBuilt on .NET 9 WPF with battle-tested libraries for every capability.
Full-featured code editor component with syntax highlighting, code folding, and line numbers.
Microsoft's Chromium-based WebView for rendering AI chat with full markdown and HTML support.
Pure .NET spreadsheet library for creating and modifying spreadsheets — no Office installation required.
Embedded document database for persistent memory storage — zero config, serverless.
Custom-built .NET HttpClient for Ollama with NDJSON streaming, tool calling, and model management — mirroring the architecture of all other provider services.
.NET speech synthesis for AI response audio playback — text-to-speech built into .NET 9.
Build in Debug for development or Release for distribution. Open in Visual Studio for the full WPF design experience.