Run every coding agent — and your IDE — in one workspace
Tejus puts Claude Code, Codex, Cursor and more side by side with a built-in IDE, a security review of every agent turn, and 1M free Tejus Cloud tokens a day.
Free · Windows 10 or 11, 64-bit (x64) · ~220 MB · macOS and Linux coming soon · SHA-256 checksum
Provider portability
Keep the runtimes you already trust.
Tejus sits around your coding agents rather than replacing them. Each CLI keeps its own authentication, models, tools, and permissions, while Tejus gives the work one thread, environment, review, and delivery layer — and the built-in Tejus agent runs on Tejus Cloud or your own API keys, all from one model picker.
The built-in Tejus agent — nothing to install. Sign in with Google for 1M free hosted-model tokens a day, or add your own API key for OpenAI, Anthropic, Gemini, DeepSeek, OpenRouter and more.
Use the Claude Code runtime and account already configured on your machine.
Run Codex tasks through its app-server with the account already signed in on your machine.
Use Cursor Agent beside other runtimes without moving the work to another workspace.
Bring OpenCode and its configured provider catalog into the same thread system.
Run Antigravity with its own account, models, permissions, and session behavior.
Use Grok Build through its local runtime and Agent Client Protocol integration.
Run Devin locally through the Agent Client Protocol with its own account and models.
Run Pi with its provider registry, thinking levels, and skills.
Run Oh My Pi through the Agent Client Protocol with the setup you already have.
Use Droid through its authenticated runtime, model controls, and ACP session.
Durable thread ownership
One workspace. Separate threads. Shared control.
A coding agent is only one part of the job. Tejus keeps the objective, environment, live processes, verification evidence, and delivery state attached to the work from first prompt to pull request.
One thread owns one line of work
Keep the objective, provider session, working directory, terminal, browser, diff, and delivery state attached to the same thread.
Providers stay portable
Use the coding-agent CLIs and accounts already on your machine next to Tejus Cloud and your own API keys — all from one model picker.
Parallel work stays isolated
Give concurrent threads separate Git worktrees and visible ownership so agents can build, test, and review without overwriting one another.
Results stay reviewable
Inspect commands, browser evidence, file changes, diffs, security findings, commits, and pull requests before accepting the result.
01 / delivery
Deliver from the same thread
Review the final diff, commit the intended changes, push, and open the pull request without reconstructing the work in another tool.
02 / parallel work
Isolate concurrent work
Give parallel threads separate Git worktrees so each agent has a clear branch, working directory, and ownership boundary.
Keep every repository ready to resume.
Move between products, client work, and experiments without rebuilding context from scattered windows. Each project keeps its own threads, provider sessions, environments, and activity state.
From objective to evidence
Keep execution and verification in the same loop.
Direct several threads, move work between providers, observe live processes, and verify the result without losing which thread owns what.
03 / thread ownership
Run parallel threads with explicit ownership
Open a thread for each objective — side by side in split view if you like — and keep its conversation, runtime, environment, and status visible instead of mixing unrelated work into one chat.
04 / provider handoff
Hand the same thread to another provider
Change the provider from the model picker while keeping the thread environment and the conversation Tejus carries forward. Review the working tree before and after every handoff.
05 / live processes
Keep long-running processes attached to the work
Run development servers, test watchers, log tails, and one-off commands beside the thread that started them — and see everything still running in the Background tasks panel.
06 / verification
Verify the behavior, not only the diff
Open the running app in the built-in browser to inspect the rendered result, check console output, and bring what you find back to the same thread.
Built into Tejus
An IDE, a security reviewer, and free tokens in the same window.
Agents do the typing, but you still read the code, check what changed, and pay for the tokens. Tejus keeps all three next to your threads.
07 / built-in IDE
Open the IDE without leaving your agents
Switch any project to the Editor view and work by hand next to the agents. Running chats keep going, and new chats started in the IDE work in its folder.
- Live file explorer — files an agent, a terminal, or another app creates appear right away
- Editor tabs, file-name search, and search in files
- Source Control with staging, diffs, commit, push, and pull requests, plus a Git history graph
- Terminals in the bottom panel and a web preview tab for your dev server
- Chat on the right that sees the open file and the lines you selected
- A familiar menu bar — File, Edit, Selection, View, Go, Terminal, Help
08 / security review
Every agent turn gets a security review
After each turn, Tejus checks the files the agent changed and tells you what it found before you commit. The review runs on your machine with local rules — nothing is uploaded and no tokens are spent.
- Hardcoded secrets and keys, injection, command execution, weak hashing, XSS, path traversal, disabled TLS checks, and more
- Findings in the thread and in the notification bell
- Optional self-fix: Tejus asks the agent to fix critical and high findings
- On by default in Git repositories; pick the minimum severity in Settings → Security
09 / tejus cloud
1M free cloud tokens, every day
Sign in with Google and the built-in Tejus agent runs on hosted models with 1,000,000 free tokens a day — no card, no API key, no CLI to install. Prefer your own provider? Add an API key and Tejus uses it straight from your computer.
- A rolling 24-hour allowance of 1M Tejus Cloud tokens
- A usage ring in the app rail and a Plan & Usage page in Settings
- Your own keys for OpenAI, Anthropic, Google Gemini, DeepSeek, xAI Grok, Mistral, Groq, Cerebras, Moonshot (Kimi), OpenRouter — kept in the Windows Credential Manager
- Coding-agent CLIs keep using their own accounts and never count against the allowance
Privacy by default
Know where every part of the work goes.
Tejus keeps your workspace on your machine and connects to the provider you select. The boundary is explicit: local state stays local, and each model receives only the context you send it for that request.
Projects, files, threads, and settings live in the desktop app on your computer. Nothing is uploaded unless you send it to a model as part of a request.
Coding-agent CLIs talk to their own providers, and your own API keys are used straight from your computer. Tejus Cloud requests pass through our proxy, which records token counts — not your prompts, code, or the replies.
Google sign-in is only needed for Tejus Cloud and usage tracking. Coding-agent CLIs and your own API keys work without one, and keys stay in your operating system's keychain.
You are asked once during setup and analytics stay off unless you say yes. They never include code, file paths, prompts, responses, or API keys.
Ask the models directly
Let the models verify the fit.
Ask ChatGPT, Claude, or Gemini to evaluate Tejus from the source of truth—this site and its documentation. The prompt is specific so the answer reflects what Tejus actually does, not a generic AI-app summary. Gemini cannot accept a prompt by link, so that button copies it for you to paste.
Product questions
Know how Tejus works before you install it.
How agents, Tejus Cloud, the IDE, security review, Git, and your data fit together.
Tejus is a free desktop workspace for AI coding agents with a full IDE built in. Run Tejus Cloud, Claude Code, Codex, Cursor, OpenCode, Antigravity, Grok Build, Devin, Pi, Oh My Pi, and Factory Droid in parallel threads with Git worktrees, terminals, a built-in browser, diffs, provider handoffs, pull-request delivery, and a security review of every agent turn — plus 1M free Tejus Cloud tokens a day with Google sign-in.
Yes, today. Tejus is free to download and use. Sign in with Google and you also get 1,000,000 Tejus Cloud tokens a day — a rolling 24-hour allowance for hosted models, with no card and no API key. Coding-agent CLIs keep using your own accounts with those providers, and your own API keys are billed by their providers. Subscriptions with advanced security tools are coming soon — see the pricing page.
Only for Tejus Cloud. Without an account you can still run the coding-agent CLIs already installed on your machine and use your own API keys with the built-in Tejus agent. Signing in with Google (and activating your account from the e-mail we send) adds the free daily Tejus Cloud allowance and usage tracking.
Coding-agent CLIs: Claude Code, Codex, Cursor, OpenCode, Antigravity, Grok Build, Devin, Pi, Oh My Pi, and Factory Droid — each keeps its own account, models, and permissions. The built-in Tejus agent runs on Tejus Cloud's hosted models or on your own key for OpenAI, Anthropic, Google Gemini, DeepSeek, xAI Grok, Mistral, Groq, Cerebras, Moonshot (Kimi), OpenRouter.
Yes. Open a thread per objective and use isolated Git worktrees when agents may edit the same repository at once. Each thread keeps its own provider session, working directory, terminal, browser, diff, and delivery state, so ownership stays visible.
Yes. Hand a thread to another provider from the model picker and the conversation carries over into the same thread environment. Review the working tree before and after a handoff, because providers do not have identical tools or session behavior.
Open any project in the Editor view (the IDE button in the title bar): file explorer, editor tabs, search, Source Control with Git history, terminals, and a web preview, with the chat on the right. The chat sees the open file and the lines you selected, and a familiar menu bar — File, Edit, Selection, View, Go, Terminal, Help — is there when you want it.
It is on by default. After every agent turn in a Git repository, Tejus checks the changed files with local rules — hardcoded secrets and keys, injection, command execution, weak hashing, unsafe deserialization, XSS, path traversal, disabled TLS verification, and more. It runs on your machine, uploads nothing, and uses no tokens. Findings appear in the thread and the notification bell; optionally Tejus asks the agent to fix critical and high findings. It never blocks your work.
Your code and files stay on your computer. Only what you send to a model as part of a request leaves it: to the coding agent's provider, to the provider your own key belongs to, or — for Tejus Cloud — through our proxy, which records token counts, not your prompts or the replies. Product analytics is opt-in: you are asked once during setup, it is off by default, and it never includes code, file paths, prompts, responses, or API keys.
Work on a normal branch or an isolated worktree, inspect the diff, commit, push, and open a pull request from the same thread. Pull requests are created through the GitHub CLI (gh).
Windows 10 and 11 (64-bit) today. macOS and Linux builds are coming soon. Tejus checks for new versions on its own and installs an update with one click.
Write to support@tejusai.com — the Tejus team reads everything.
The desktop workspace for coding agents, with a full IDE built in.
Run every coding agent — and your IDE — in one workspace
Start with one repository and one agent. Add parallel threads, worktrees, the IDE, browser checks, handoffs, and pull-request delivery when the work needs them.
Free · Tejus 1.0.3 · Windows 10 or 11, 64-bit (x64) · ~220 MB · macOS and Linux coming soon