Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

724 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Tengra Logo

Tengra

Cross-platform AI workspace for developers and everyday users.

License Platform Build and Test Build and Release



Interface

Home Page
Chat, templates, recent conversations, and the main composer in one view.


Workspaces
Workspaces for project-based organization and deeper task flows.


Image Studio
Integrated image studio for prompt-driven generation inside the app.


Marketplace
Marketplace surface for modules, themes, models, and other add-ons.


Settings
Settings for app behavior, accounts, models, and customization.


What It Covers

  • Chat and model switching: Use hosted models or local models from one desktop surface.
  • Workspace control: Work directly in local files and Git-backed projects.
  • Image generation: Generate images without leaving the app.
  • Portable mode: Keep app data inside the application folder for a zero-install flow.
  • Native sidecar services: Use the Rust-backed path for faster and more isolated system operations.
  • Everyday AI tasks: Draft, summarize, debug, translate, and generate content without switching apps.

The Landscape

Product Pricing / access Platforms Open source Local models Image gen Agent surface Workspace / file control Privacy posture Extensions / MCP / plugins Best for Not ideal for
Tengra Open source Windows, macOS, Linux Yes Yes Yes Chat, workspace, image Edit files, terminal, Git Local models and portable mode Plugins and MCP Native desktop AI with local models Browser-only users
ChatGPT / Codex Subscription Desktop, web, terminal Partial, Codex CLI open source No Yes in ChatGPT Chat and coding agent Read, write, run Cloud-hosted Limited native plugin story General AI and coding Strict local-first setups
Claude Code Subscription Terminal, web No No No CLI agent Read, write, run Cloud-hosted MCP Terminal-first coding agent GUI-heavy workflows
Antigravity Preview / quota Desktop, browser No No core docs No Editor, browser, mission control Plans, edits, validates Google account and preview quotas Browser and integrations Google-first agentic workflows Local-first or open-tooling users
Windsurf Free and paid Desktop No No No core image-gen feature Chat, command, workflows Edit files, terminal, deploy Cloud account MCP and Open VSX Agentic IDE workflows Open-source purists
Cursor Subscription Desktop, terminal No No No Editor and background agents Edit files, terminal Privacy mode available MCP AI-first IDE Open-source purists
PearAI Open source Desktop Yes Partial No Editor and chat File editing Local-friendly, but docs are thin VS Code ecosystem Open-source VS Code fork People who want mature official docs
OpenClaw Open source Desktop, self-hosted Yes Yes Partial Multi-channel assistant gateway Local files and tools Local-first Skills and channels Local-first assistant gateway Simple single-app chat
Continue Open source IDE, CLI Yes Yes No Chat, edit, apply, autocomplete Edit, apply, terminal Local or BYOK MCP and IDE integrations Extend existing IDEs Standalone desktop app
Zed Free and paid Desktop Yes Yes No Agentic editor Edit selections and external agents Local-first ACP and external agents Fast native editor with AI Chat-centric non-editor users
VS Code Free Desktop, web Partial, repo open source Yes via tooling No Local and background agents Edit, run, Git Depends on extensions and models Extensions and MCP General-purpose editor Opinionated AI workflows out of box
Copilot CLI Subscription Terminal No No No Terminal agent Read, modify, run Cloud-backed GitHub integration Terminal-only coding tasks Desktop-first workflows

Verified against official docs on 2026-05-16.


Tengra sits between a general-purpose desktop assistant and a developer-first workspace: GUI-friendly for everyday AI use, but built for managed workspaces, local models, and a native engine when the task needs more control.


Getting Started

  1. Download a pre-built package for your platform from the releases page.
  2. Launch Tengra and sign in or configure local models.
  3. Open a workspace, switch models, and start chatting or editing.

Project Docs


Technologies


Core Capabilities

  • Hybrid Inference: Switch between OpenAI/Claude and local Ollama/LM Studio models instantly.
  • Managed Workspaces: Native Git integration and direct filesystem project management.
  • Native Sidecar: Rust-based engine for high-speed file I/O and secure proxying.
  • Portable Mode: Zero-install operation with 100% data redirection to the application folder.

Installation

Pre-built

Development

For Windows developers, the Rust proxy requires llvm and nasm to compile the Chrome-impersonation engine (BoringSSL). You can install them via Scoop or Chocolatey:

# Using scoop
scoop install llvm nasm

# OR using chocolatey
choco install llvm nasm -y
git clone https://github.com/TengraStudio/tengra.git
npm install
npm run dev

Star History

Star History Chart


License

Tengra is licensed under GPL-3.0. See LICENSE for details.


Made with ❤️ by TengraStudio

About

Cross-platform AI workspace for developers — local models, Git integration, Rust-powered sidecar.

Topics

Resources

Code of conduct

Contributing

Security policy

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages