Kortix vs Nimbalyst: an open-source company system or a coding-agent workspace?

Canonical: https://kortix.com/blog/kortix-vs-nimbalyst

Nimbalyst is an open-source visual workspace for the coding agents you already run, and Kortix is the open-source system that runs the agents and the rest of a company's work from one git repo you own. If your job is steering Claude Code and Codex and reviewing their diffs, Nimbalyst is built for exactly that. If your job is owning the agents, the memory, and the connectors across a company, the answer is Kortix.

What Nimbalyst is good at

Nimbalyst is an open-source desktop workspace that runs Claude Code, OpenAI Codex, and other coding agents on your own machine, with visual editors for the work around the code (Nimbalyst, checked October 2026). It keeps the agents you already pay for and wraps a working surface around them.

• Parallel coding sessions, each in its own worktree. Run several Claude Code and Codex sessions at once, optionally isolated in git worktrees so they do not collide, and track them on a kanban board by phase.
• Visual editors for the whole project. WYSIWYG markdown, Excalidraw and Mermaid diagrams, mockups, data models, spreadsheets, CSV, and code in Monaco, all in the window the agents edit.
• Inline red and green diff review. Agent edits are written to disk and shown as inline diffs in the editor, so you accept or reject each change before you commit (Nimbalyst features).
• The agents you already use. Claude Code and Codex are first-class; OpenCode and GitHub Copilot run as alpha providers, and an extension can register another. You sign in with the Claude or ChatGPT subscription you pay for.
• Free and MIT licensed for individuals. The desktop and iOS apps are MIT licensed and free for individual use with no feature limits, on macOS, Windows, and Linux (Nimbalyst pricing).
• A local-first file model. Files stay on your disk in their normal formats, so the same repository works with or without Nimbalyst.

For one developer directing coding agents, and for a small team that wants that in a shared workspace, Nimbalyst is a well-built tool. It is SOC 2 Type 2 certified, runs its collaboration server inside the customer's own Cloudflare account, and sells a team edition at $20 per user per month, free while the beta lasts (Nimbalyst Enterprise, pricing, checked October 2026). The desktop and iOS apps are MIT licensed, and the server is a separate restricted-license service.

Where a coding-agent workspace stops

Nimbalyst is built around the coding agent. It runs the vendor's own agent software on your machine and adds the editors, review, and board around it. That focus is the point of Nimbalyst, and it sets the boundary.

• The job is code. Nimbalyst is a coding-agent workspace. It does not run outbound, reconcile the books, or triage support across departments.
• State lives per machine, not in a company repo. Sessions, tasks, and a context graph persist, and the files are open markdown. No single repository holds a company's agents, skills, memory, and connector rules as versioned files a team reviews together.
• Execution is local. The agent runs on each person's machine using local files and that person's subscription. That suits a developer steering an agent, and it is not one isolated cloud computer per session on a shared config.
• The workspace does not run the agents. Nimbalyst manages Claude Code and Codex; it does not host the compute they run on, start them on a schedule with nobody present, or put every department's work behind one review gate.
• Reach is developer-shaped. The tool surface is git, pull requests, MCP servers, and team documents rather than the thousands of business apps a company-wide workforce touches.

Nimbalyst does that job well, and these are the limits of the job itself.

What Kortix does differently

Kortix is the open-source AI Operating System: one git repository holds the agents, skills, company memory, connector configuration, and triggers, and every session runs on its own isolated cloud computer (Kortix docs). The company is the repository.

A short kortix.yaml declares how a session boots:

# kortix.yaml (excerpt; the manifest lives at the repo root)
default_agent: ops-morning
sandbox:
  image: Dockerfile
agents:
  ops-morning: { connectors: [slack, stripe] }
triggers:
  - { type: cron, agent: ops-morning, cron: '0 7 * * 1-5' }

• One repo you own. Agents and skills are Markdown, memory is files, and connector rules and triggers are config. Grep the whole company, diff any change, roll it back.
• One gate to land work. A session runs on a branch named after it, the agent commits, and the work reaches main only through a change request a person reads as a diff. Merge is default-deny for agents.
• An isolated machine per session. Every session boots its own sandbox on its own branch. Thousands run in parallel on the same config with no crossover between them.
• Any model, your keys. Pick a model per agent, per session, or per message: any provider with your own keys, or the ChatGPT plan you already pay for. OpenCode is the harness.
• Every business tool. Kortix connects 3,000+ apps in a click, plus MCP, OpenAPI, Postman, GraphQL, and raw HTTP. Connector credentials are brokered server-side and never enter the machine.
• Deploy anywhere. Self-host on a laptop, a VPS, your VPC, or on-prem, or use managed Kortix Cloud. Self-host is free; Free is $0 with 200 credits each month and 1 project, and the team plan is $40 per seat per month with 2,500 pooled credits per seat (Kortix pricing, checked October 2026).

Kortix is open source: read it, fork it, audit it, and self-host for free. The source is at Kortix on GitHub.

Where Nimbalyst wins

For a developer or a small team that wants to steer coding agents visually, Nimbalyst is the better start. It runs the exact agents many teams already pay for, keeps files on the local disk, and gives diff review and parallel sessions that a terminal alone lacks. The MIT license and the free individual tier remove the cost question. A team whose entire job is code, and that wants a local workspace around it, should start there. Kortix asks you to treat the whole company as a repository, which is more than an engineering team needs.

Side by side

When to pick which

For a coding-agent control center on your own repositories, read the Kortix vs OpenHands comparison. To choose a platform rather than a workspace, see how to pick an open-source AI agent platform.

Run the whole company from one repo you own. Kortix is free to start and free to self-host, and the agents, skills, memory and connectors stay in a git repo your team keeps. Get started with Kortix.