gonuUnderstand the system
Explore repository structure and the context behind a change.
Explore the BrainMeet your Engineering BrainBetaYour AI
FOR DEVELOPERS & ENGINEERING TEAMS
Your AI
Engineering Team.
Build, debug and review with the context that matters.
Your code. Your team’s knowledge. Your control.
Start without signing in · No personal API key needed
YouAdd a retry limit to the checkout request.
A smaller change. A clearer plan.
Find the request path, preserve the existing behavior and propose a focused change.
- Read the checkout request and its callers
- Bound retries without changing the public API
- Add a test for the retry limit
Sample screens · no live executionExplore all demos
Good work starts where you are.
Compare all toolsWeb Chat
Try nowDesktop Worker
DownloadVS Code Extension
Source betaGonu CLI
Source betaGonu IDE
Not releasedOne Gonu account. Your saved agents and task history across supported clients. How shared access works
MORE CONTEXT. LESS STARTING OVER.
Software has a history.
Give it a memory.
A file only tells part of the story. Bring repository structure, recorded decisions and selected team knowledge into the same view.
Explore the Engineering Brain Engineering Brain / sample-checkoutIllustrative map · sample data
START FROM THE CODEUnderstand what connects.
Checkout servicesrc/api/checkout.ts
Request pathrouter → checkout → paymentsCode
Why retries are boundedSelected architecture decisionDecision
What needs a checkTimeout and duplicate-order casesContext
Keep the reasons
Revisit recorded decisions and engineering history instead of starting from zero.
See Software Time MachineLook for the evidence
Inspect recorded changes and checks, including what still needs verification.
See Proof of WorkONE PLATFORM. REAL WORK.
From your next fix
to your team’s next step.
FOR DEVELOPERS
Stay close to the code.
Open a repository, understand the execution path and work toward a focused, reviewable change.
- Ask and plan before editing
- Inspect diffs, checks and task history
- Use project rules and your selected agents
CLI · source beta
$ gonu plan "Explain the checkout flow"Requires CLI setup and an authorized project.FOR BUSINESSES & TEAMS
Make context a team habit.
Turn selected documents and meeting transcripts into source-linked drafts, handoffs and next actions.
- Keep shared project knowledge in reach
- Review meeting decisions and action items
- Organize work with workspace roles
CAPABLE AI. CLEAR BOUNDARIES.
The work moves forward.
The decisions stay yours.
Gonu provides the models. You choose the context and review consequential actions.
Models, usage & control- 01
Bring the right context
Start with a question, selected documents or an authorized repository. Connect what the task needs.
- 02
Make the next step clear
Use Auto to get started. Ask for a plan, choose your agents and inspect proposed changes.
- 03
Review before moving on
Check the output, diffs and execution evidence. Keep approvals in place for consequential actions.
Hosted models process selected context with the provider. Worker also supports optional local Ollama inference.
Understand your data flowSTART WITH ONE USEFUL RESULT
Find your flow.
Then choose your plan.
Try the core experience first. Check current prices, included usage and available capabilities before upgrading.
Start free
Try web chat. Sign in for saved work, your agents and Auto within your free allowance.
Try Gonu on webMake it your workflow
Compare subscriptions for model selection and additional usage. Limits and availability apply.
Compare plansStart a team pilot
Bring a real use case. Discuss your repositories, team setup and model requirements.
Request a team demoBEFORE YOU START
A little clarity.
A better first step.
Practical answers, setup guides
and a closer look at the product.
Can I try Gonu without signing in?
Yes. Open Web Chat for a limited guest conversation, with no download or personal API key. Sign in to save conversations and use your agents. Guest chat uses Auto and keeps model and agent controls out of the way.
Can web chat change files on my computer?
Web chat cannot open local folders or execute commands on your computer. Use Desktop Worker, the VS Code beta or CLI beta for authorized local repository work. Managed engineering workflows need a connected project and the required permissions.
Which models can I use?
Gonu provides hosted models. Free accounts use Auto; subscriptions unlock model selection, subject to available models, usage limits and workspace policies. Optional local Ollama inference is available in Worker. You do not need to bring a personal API key for Gonu-hosted models.
What is the Engineering Brain?
It connects repository structure with recorded decisions, engineering history and selected team knowledge. The current beta starts from an authorized repository scan and the sources you choose. Each feature page explains the available workflow, prerequisites and what is still being built.
Are the screens on this page real executions?
The interactive workflow and repository map are clearly labeled, authored examples. They explain how the product works without running commands or calling a model. The workspace image is an actual product screen with example data. A real task needs your inputs and setup; its results must be verified.
Is my code kept on my device?
That depends on your setup. Hosted models process the selected context with the provider. Local Ollama runs inference on your computer, while account and some workflow features still use Gonu services. Review the model and data guide before choosing a setup for sensitive work.