Open Interpreter
Open-source code agent now expanding into a desktop app
Overview
Open Interpreter runs LLM-written code locally via natural language. The classic Python CLI now lives on as a community fork; the official project has moved to a Rust rewrite plus a new desktop app, Interpreter, that edits Word/Excel/PDF files and controls your computer directly.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 20 Jul 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- Bring your own API key
- Local models via Ollama/LM Studio/Llamafile
- Sign in with OpenAI for ChatGPT access
- Community support
- Hosted model access
- ~4x usage vs free tier
- Desktop document/PDF/Excel agents
- ~3x usage over Pro
- Team seats
- Hosted models included
Use cases
What you can produce with Open Interpreter
- Natural-language-to-code execution (Python/shell/AppleScript)
- Local model support via Ollama, LM Studio, or Llamafile
- Rust-based CLI tuned for open models (DeepSeek, Kimi, Qwen)
- Interpreter desktop app: Word, Excel, and PDF editing agents
- Approval-gated command execution (manual confirm per action)
- Sign in with OpenAI for free ChatGPT-based use
ASEAN Perspective
Open Interpreter in Southeast Asia
ASEAN-region availability and pricing notes coming soon. Drop the editorial team a note via /contact/ if you can supply local context (Singapore/Malaysia/Indonesia/Thailand/Vietnam).
The product behind this entry has moved twice since it built its reputation. The original Python CLI — the thing that earned 50,000+ GitHub stars letting an LLM write and run Python, shell, and AppleScript on your machine — is no longer where the official org's attention goes; that codebase now survives as a community-maintained fork (endolith/open-interpreter). The official repository has been repointed to a from-scratch Rust rewrite built for open models like DeepSeek, Kimi, and Qwen, and creator Killian Lucas has launched a separate polished desktop app, simply called Interpreter, that expands the pitch from "run code" to "operate your computer" — editing Word docs, building Excel pivot tables, filling PDF forms.
The core risk hasn't changed: this executes real commands on a real machine with your permissions, and the project's own docs still recommend a container. Approval-gated execution (press Y before anything runs) is the main safeguard. Free and open, with an optional $20-60/month hosted plan for the desktop app.
What people say
This entry no longer describes one project — it describes two, and neither matches the original stub's framing. The GitHub org's flagship repository has been repointed: it now hosts a ground-up Rust rewrite (64,300+ stars as of writing, latest release 0.0.21 shipped 6 July 2026) described in its own README as "a lightweight coding agent for open models like Deepseek, Kimi, and Qwen." The original Python codebase — the one most existing comparisons and reviews still describe — lives on only as a community-maintained fork at endolith/open-interpreter, with its own maintainer warning the active branch is "a vibe-coded mess, use with caution."
Alongside the CLI rewrite, creator Killian Lucas has shipped a separate consumer-facing product: a desktop app called simply Interpreter, positioned as a general-purpose computer agent rather than a coding tool. It adds document editing the original CLI never had — filling non-interactive PDF forms, building pivot tables and charts in Excel, editing Word docs with tracked changes — and supports signing in with an OpenAI account for free ChatGPT-based use, plugging in Anthropic/OpenAI/Groq/OpenRouter keys, or running fully local models via Ollama.
The security framing hasn't softened. The project's own safety docs still warn that giving an LLM shell access on a personal machine is a real risk and recommend containerizing it; unlike sandboxed tools such as ChatGPT's Code Interpreter, this runs with the user's actual system permissions, so a bad or hallucinated command can delete files or change configuration. The interactive Y-to-approve step before execution is the main guardrail, and reviewers note most users skim past the warnings and only go looking for a safer alternative after something breaks. An OpenSSF Scorecard of 5.0/10 reflects mixed supply-chain hygiene (branch protection, signed releases, dependency review) rather than any specific incident.
Pricing has also shifted from purely free to a freemium desktop product: base use is free with your own API key or local models, a $20/month Pro plan adds hosted models and roughly 4x the usage, and a $60/month Business plan adds team seats at 3x Pro's allowance.
Summary of public user & expert reviews, compiled by RECATOOLS.
About this listing
This entry was compiled from publicly available data including Open Interpreter's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Open Interpreter unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Open Interpreter directly →
Spotted something out of date? Suggest an update →
Open Interpreter in the news
More in Code & Dev Tools
