TT-VSCode-Toolkit

Interactive AI lessons, hardware monitoring, and production inference templates — all inside your editor. Works on Wormhole™ (n150, n300, T3000) and Blackhole® (p150, p300, TT-QuietBox® 2) hardware.

Install from VS Code Marketplace Download .vsix Browse Lessons — no hardware needed →

Or via terminal: code --install-extension Tenstorrent.tt-vscode-toolkit

Extension guide → GitHub →
TT-VSCode-Toolkit — lesson panel inside VS Code showing interactive lessons with hardware badges and one-click terminal commands

Built-in features

Everything You Need to Ship on Real AI Hardware

More than documentation — the toolkit runs shell commands for you, monitors your hardware, and manages inference servers, all without leaving your editor.

📖

Interactive Lessons

One-click buttons that run the exact terminal commands described in the text. No copy-pasting, no tab-switching — the editor does the work.

🔧

Hardware Filter

Select your device — n150, n300, T3000, p100, TT-QuietBox 2 — and the lesson panel narrows to only lessons validated on your hardware. Filter state persists across sessions.

🚀

Live Server Control

Start and stop vLLM, TT-Inference-Server, and the prompt generator directly from the extension. Logs stream into a collapsible panel until the health check clears.

⌨️

70+ Commands

Every common operation — hardware detection, model download, environment setup, compiler activation — wired into the Command Palette (Ctrl+Shift+P).

What you can build

Generated on Tenstorrent Hardware

Sample outputs from the Cookbook lessons and applications — all running on the Wormhole and Blackhole chips the extension targets. Agent-generated art from the Local AI Agents lesson runs on TT-QuietBox 2.

Tenstorrent hardware in action
TT-VSCode-Toolkit walkthrough
AnimateDiff / Wan2.2 — video generation
AnimateDiff on Blackhole — World of Tomorrow, 8 frames × 25 denoising steps on p300c AnimateDiff on Blackhole — World of Tomorrow (p300c)
AnimateDiff on Blackhole — Phosphor Horizon, 8 frames × 25 denoising steps on p300c AnimateDiff on Blackhole — Phosphor Horizon (p300c)
Stable Diffusion XL image generation — snowy cabin Stable Diffusion XL — image generation
Conway's Game of Life running on TT cores — Recipe 1 Recipe 1 — Conway's Game of Life
Particle Life emergent behavior simulation across multiple TT devices Recipe 5 — Particle Life (multi-device)
Mandelbrot fractal explorer — Recipe 3 Recipe 3 — Mandelbrot Explorer
Julia set fractal Julia Set variant
Mel spectrogram from Recipe 2 — Audio Signal Processing on TT hardware Recipe 2 — Mel Spectrogram
Generative landscape SVG created by a local AI agent on TT-QuietBox 2 Local agents — generative landscape

Lesson library

What You'll Learn

From your first inference run to deploying production-grade inference servers — each track builds on the last.

🚀

Your First Inference

  • Modern Setup with TT-Installer 2.0
  • Hardware Detection
  • Download Model & Run Inference
🏭

Serving & APIs

  • Production Inference with TT-Inference-Server
  • Production Inference with vLLM
  • Image Generation with Stable Diffusion XL
🎬

Applications

  • Native Video Animation with AnimateDiff
  • OpenClaw AI Assistant on TT-QuietBox 2
  • Generating Video on TT-QuietBox 2
🍳

Cookbook

  • Recipe 1: Conway's Game of Life
  • Recipe 3: Mandelbrot Fractal Explorer
  • Recipe 5: Particle Life Simulator
🔧

Compilers & Frameworks

  • Image Classification with TT-Forge™
  • JAX and PyTorch/XLA on Tenstorrent
🧠

Advanced

  • Exploring TT-Metalium™
  • Bounty Program: Model Bring-Up
🎓

CS Fundamentals

  • RISC-V & Computer Architecture
  • The Memory Hierarchy
  • Parallel Computing
⚙️

Custom Training

  • Understanding Custom Training
  • Dataset Fundamentals
  • Configuration Patterns

Browse all lessons →

Get started

Install in Your Editor

The easiest way is the VS Code Marketplace — one click or one command. A .vsix is also available for Cursor, Windsurf, and code-server.

1

Install directly from the VS Code Marketplace with one command:

code --install-extension Tenstorrent.tt-vscode-toolkit
2

Or open VS Code, press Ctrl+Shift+X, search for TT-VSCode-Toolkit (extension ID: Tenstorrent.tt-vscode-toolkit), and click Install.

Open in Marketplace →

3

Reload VS Code — the TT-Toolkit icon will appear in the Activity Bar. Click it to open the Lessons panel and pick your hardware.

1

Download the .vsix from the latest GitHub release.

gh release download --repo tenstorrent/tt-vscode-toolkit --pattern '*.vsix'
2

Install via the terminal:

code --install-extension tt-vscode-toolkit-*.vsix
3

Reload VS Code — the TT Toolkit icon will appear in the Activity Bar. Click it to open the Lessons panel and pick your hardware.

1

Download the .vsix from the latest release.

gh release download --repo tenstorrent/tt-vscode-toolkit --pattern '*.vsix'
2

Open the Extensions panel (Ctrl+Shift+X), click the ⋯ menu, choose Install from VSIX… and select the downloaded file.

Both Cursor and Windsurf are compatible with standard .vsix packages.

1

Download and install in one command:

gh release download --repo tenstorrent/tt-vscode-toolkit --pattern '*.vsix' \
  && code-server --install-extension tt-vscode-toolkit-*.vsix
2

Or via the GUI: open the Extensions panel (Ctrl+Shift+X), click ⋯ → Install from VSIX….

1

One command downloads and installs the latest release (requires GitHub CLI): 

gh release download --repo tenstorrent/tt-vscode-toolkit --pattern '*.vsix' \
  && code --install-extension tt-vscode-toolkit-*.vsix
2

Re-running the same command updates to the latest version (use --clobber to overwrite the existing .vsix).