Open Interpreter · 2026-07-16 · major
Open Interpreter 0.0.26 — Rust coding agent tuned for open models
Open Interpreter's Rust rewrite is out: a Codex-compatible coding agent for low-cost open models. Three tagged releases shipped in 72h; the terminal UI swaps between Claude Code, Kimi CLI, Qwen Code, and DeepSeek TUI harnesses.
The 66k-star coding agent has been rewritten in Rust to make cheap open models act more like Codex.
Key specs
| GitHub stars | 65.9k |
|---|---|
| Stars today | 633 |
Quick facts
| Maker | Open Interpreter |
|---|---|
| Version | 0.0.26 (Rust) |
| License | Apache-2.0 |
| Language | Rust |
| Model support | claude-code, kimi-cli, qwen-code, deepseek-tui |
| Platforms | macOS, Linux, Windows |
| Install | curl openinterpreter.com/install | sh |
What is it?
Open Interpreter 0.0.26 is the first stable tag of the project's Rust rewrite. The tool now ships as a single interpreter binary with native sandboxing and swaps between four built-in agent harnesses — claude-code, kimi-cli, qwen-code, and deepseek-tui — from the /model command inside a terminal UI.
How does it work?
Under the hood the new build mirrors OpenAI's Codex agent loop, then wraps it around the tool-use format each backend understands, which is what lets a smaller open model imitate the planning and tool discipline of a frontier one. Config lives in ~/.openinterpreter, and the same binary can act as an Agent Client Protocol server for editors.
Why does it matter?
A serious coding agent aimed at open models matters because Kimi K3, Qwen, and DeepSeek are cheap or free but usually feel weaker than Claude or GPT inside off-the-shelf agents. Open Interpreter's rewrite closes that gap by giving those models the same harness the paid frontier tools use, which is why the repo picked up 633 stars in a single day.
Who is it for?
developers running local or low-cost open models
Frequently asked questions
- What is Open Interpreter 0.0.26?
- Open Interpreter 0.0.26 is the new Rust build of the popular open-source coding agent, based on the Codex architecture and tuned to squeeze strong tool-use behavior out of cheaper open models like Kimi K3, Qwen, and DeepSeek. It ships as a single terminal binary with native sandboxing on macOS, Linux, and Windows.
- Why did Open Interpreter rewrite in Rust?
- The Rust rewrite lets Open Interpreter emulate the same agent harness Codex uses, so a small open model can act more like a frontier one. The original Python project continues as a community fork, but the Rust build is the new default and is the version being tagged in the 0.0.24 to 0.0.26 releases from July 14 to July 16.
- Which models does Open Interpreter 0.0.26 support?
- Open Interpreter 0.0.26 ships four built-in harnesses that can be swapped with the /model command: claude-code for Anthropic, kimi-cli for Moonshot's Kimi K3, qwen-code for Alibaba's Qwen, and deepseek-tui for DeepSeek. Any Codex-compatible provider works through the same interface.
- How do you install Open Interpreter 0.0.26?
- On macOS or Linux, run curl -fsSL https://www.openinterpreter.com/install | sh; on Windows, use the PowerShell one-liner from openinterpreter.com/install.ps1. The installer drops a single interpreter binary and stores config under ~/.openinterpreter.
- Is Open Interpreter free to use?
- Open Interpreter is free and open source under the Apache-2.0 license, and the tool itself has no subscription. You only pay for whichever model provider you plug in via the harness; open weights like Kimi K3 or DeepSeek can be run locally at no per-token cost.
Try it
curl -fsSL https://www.openinterpreter.com/install | sh