# Ipek Review

A local-first desktop IDE for founders that turns AI-written workflows into scheduled, inspectable data operations on your own machine.

## Editor's Score: 62/100

Capability lands at 17 because the local execution model, per-run inspectable folders, SQLite state persistence, and screen-to-context recording are genuinely differentiated features, but the tool is at v1.1 with no team collaboration, no cloud fallback for missed schedules, and a build path that requires a separately configured MCP client before anything works. Ease of use scores 13 because the entire onboarding assumes you already have Claude Code, Codex, or Cursor set up with API keys; a founder who does not will hit a hard wall before seeing a single workflow run. Value scores 22 because the tool is completely free with no per-task fees and no account required, which is genuinely strong for its target user, even accounting for the bring-your-own-AI cost. Delivery scores 10 because there is no user sentiment, no third-party coverage, and the terms of service explicitly warn that the product is free for now with no SLA or support commitment; additionally, AI-generated worker code runs with full OS user account privileges, which is a real trust and safety consideration the terms flag but do not mitigate.

## Pros
- Fully free with no per-task fees and no account required; you pay only for the AI and API keys you already use.
- Complete local execution: every project file, run output, and API response is stored on your own machine, with no backend server receiving your data.
- Per-run output folders with saved API responses make debugging and rerunning from any step concrete and practical, not a log-stream guessing game.

## Cons
- Schedules only run while Ipek is open and the machine is awake, meaning any workflow that needs to fire overnight or on an unattended machine will silently miss its window.
- The entire build path depends on having a configured MCP-compatible AI client (Claude Code, Codex, or Cursor); there is no built-in editor or manual workflow builder for users who do not have one set up.
- No paid tier exists yet, which means no SLA, no support commitment, and no roadmap guarantee; the terms of service explicitly warn that the product is provided free of charge for now, with paid plans possible later.

## Pricing

Free.
- Free: $0

Pricing reflects what we saw at the time of review (2026-11). Confirm current pricing on the official site.

Category: AI Productivity Tools (https://aitoolseekers.com/category/ai-productivity-tools)
Official site: https://ipek.run/
Last verified by a human: 2026-11-04

Reviewed by a human at AI Tool Seekers, a hand-reviewed AI tools directory. Every listed tool is verified by a person before publishing. Scores are editorial opinions from a fixed rubric (capability, ease of use, value, delivery on promise) and cannot be bought; there is no pay-to-rank.
Directory: https://aitoolseekers.com
This review: https://aitoolseekers.com/tools/ipek
Methodology: https://aitoolseekers.com/how-verification-works

## Verdict

Ipek is the right pick for a solo technical founder who wants AI-built data pipelines that run entirely on their own machine with no per-task fees and no cloud vendor holding their data. Its bring-your-own-AI model (Claude Code, Codex, Cursor) means you pay only for the AI you already use, but that also means setup requires you to have and configure an MCP-compatible client before you get anything running. Teams expecting a polished SaaS experience or non-technical users who cannot configure API keys should look elsewhere.

Best for: Solo founders automating data ops locally with their own AI client
Not for: Non-technical users or teams wanting a hosted, no-config SaaS automation tool

## Overview

Ipek is a desktop application that positions itself as an "operations IDE" for startup founders. The core idea is straightforward: your AI client (Claude Code, Codex, Cursor, or any MCP-compatible tool) connects to Ipek over localhost, then plans, writes, and wires together small programs called workers into a workflow on a visual canvas. You observe the resulting dashboard and ask your AI for changes. Nothing runs on Ipek's servers; every project, run log, and output file stays on your computer.

What separates Ipek from cloud-based automation tools like Zapier or Make is the execution model. Workers are compiled Kotlin code your AI writes, not drag-and-drop action blocks. Each run gets its own folder with every step's output and API response saved as inspectable files, so debugging means opening a folder rather than reading a log stream. A built-in SQLite database (the Project DB) lets workflows remember state between runs, such as which tickets were already grouped last time. Schedules run locally on a fixed cadence: every 10 minutes, weekdays at 09:00, or once a week, as long as Ipek is open and the machine is awake.

The tool pivoted from an earlier product called LimanDoc, and v1.1 is the current release. That early-stage reality shows in the feature set: the MCP integration is the primary build path, screen recording for context capture is a notable differentiator, but there is no cloud sync, no team collaboration layer, and no mobile access. The bring-your-own-AI and bring-your-own-API-key model means Ipek itself costs nothing to run, but you are assembling your own stack.

## Who should use it

Ipek fits a specific profile: a technical solo founder or small-team operator who already pays for an AI coding assistant (Claude Code, Codex, or Cursor), wants to automate data pulls from APIs like Linear or Zendesk, and cares enough about data privacy to prefer local execution over a hosted service. The screen-recording-to-context feature is genuinely useful for someone who wants to show their AI what to automate rather than describe it in text. If you are already running an MCP client and want scheduled, inspectable pipelines without a monthly SaaS bill, Ipek is worth the setup time.

If you need a no-code tool your whole team can use without touching API keys or a terminal, Ipek is not ready for you. Notion AI is a better fit for teams that want AI assistance embedded in a shared workspace they already use daily. For pure file and data utility work that does not require scheduling or AI-written code, EveryTool4U covers many browser-based tasks with zero setup.

## FAQ

Q: Does Ipek cost anything to use?
A: Ipek is currently free with no account required and no per-task fees. The terms of service state that paid plans may be introduced in the future, but no pricing tiers exist as of November 2026.

Q: Do I need to provide my own AI subscription to use Ipek?
A: Yes. Ipek connects to your own AI client (Claude Code, Codex, Cursor, or any MCP-compatible tool) over localhost. You supply your own API keys for AI providers like Anthropic or Google. Ipek itself does not include an AI model.

Q: Where does my data go when I run a workflow?
A: Everything stays on your machine. Ipek operates no backend server that receives or stores your projects, inputs, or outputs. API responses and run outputs are saved as files in local folders on your computer.

Q: What happens if my computer is asleep when a scheduled workflow is supposed to run?
A: The workflow will not run. Schedules execute locally only while Ipek is open and the machine is awake. There is no cloud fallback to trigger missed runs.

Q: Which operating systems does Ipek support?
A: Ipek v1.1 supports macOS (Apple Silicon and Intel) and Windows 10 and 11. There is no Linux build, web app, or mobile version.

Q: Can I edit the worker code Ipek's AI writes?
A: Yes. Workers are plain compiled Kotlin files you can open and read. Parameters like date ranges can be changed without touching the code, and the full source is accessible for manual edits.