#4040th of 70 in Agent frameworks and SDKs

mobilerun

LLM agent framework that taps, swipes and types on Android and iOS devices

Stars
9.7k
License
MIT
Last commit
Oct 2026
Last release
Oct 2026
Language
Python
Stars, last 29 weeks+1,783
1 Apr 2026: 7,946 stars8 Apr 2026: 8,012 stars15 Apr 2026: 8,061 stars22 Apr 2026: 8,103 stars29 Apr 2026: 8,140 stars6 May 2026: 8,186 stars13 May 2026: 8,221 stars20 May 2026: 8,277 stars27 May 2026: 8,335 stars3 Jun 2026: 8,401 stars10 Jun 2026: 8,451 stars17 Jun 2026: 8,512 stars24 Jun 2026: 8,551 stars1 Jul 2026: 8,608 stars8 Jul 2026: 8,649 stars15 Jul 2026: 8,720 stars22 Jul 2026: 8,787 stars29 Jul 2026: 8,839 stars5 Aug 2026: 8,942 stars12 Aug 2026: 8,993 stars19 Aug 2026: 9,044 stars26 Aug 2026: 9,090 stars2 Sept 2026: 9,131 stars9 Sept 2026: 9,305 stars16 Sept 2026: 9,353 stars23 Sept 2026: 9,421 stars30 Sept 2026: 9,526 stars7 Oct 2026: 9,574 stars14 Oct 2026: 9,650 stars21 Oct 2026: 9,729 stars

7,946 on 1 Apr 20269,729 on 21 Oct 2026

From GitHub's daily star counts.

Checked on GitHub.

Overview

Mobilerun drives Android and iOS devices from natural-language tasks through a CLI, terminal UI or Python API. A Portal app on the device exposes the accessibility tree, screenshots and input actions to the LLM agent, with an optional manager-executor reasoning mode for longer tasks. It works with OpenAI, Anthropic, Gemini, xAI, Ollama, DeepSeek, OpenRouter and OpenAI-compatible endpoints.

Who it is for: Developers automating or testing mobile apps with LLM agents

Strengths

  • Combines accessibility tree with screenshots; --vision-only covers apps without a11y data
  • Many providers supported, including local Ollama models
  • Custom tools, structured output and app cards extend agent behavior in Python
  • Tracing via Arize Phoenix or Langfuse, plus saved trajectories

Weaknesses

  • Requires ADB, USB debugging and installing the Portal app on the device
  • iOS setup is a separate flow and is documented only briefly
  • Python 3.14 is not supported; needs >=3.11,<3.14
  • Hosted devices and fleet features are in the paid Mobilerun Cloud, not the framework

What it needs

  • no GPU
  • Docker
  • Needs ADB, Mobilerun Portal app, Android device or iOS device/simulator, LLM provider API or Ollama
  • Models: OpenAI, Anthropic, Gemini, xAI, Ollama

Also in Agent frameworks and SDKs

See all 70
Also in Agent frameworks and SDKs
RankProjectScore
1ppt-masterAgent skill that turns documents into natively editable PowerPoint decks Live demo ↗ (opens in a new tab)65.3k starsMIT94 out of 100
2langchainPython framework for building LLM agents and apps from interoperable components148.2k starsMIT89 out of 100
3browser-usePython library that lets an LLM agent control a web browser Live demo ↗ (opens in a new tab)118.8k starsMIT89 out of 100
#4openai-agents-pythonPython SDK for multi-agent workflows with handoffs, guardrails and tracing30.2k starsMIT89 out of 100
#5mastraTypeScript framework for building AI agents and workflows29k starscustom license89 out of 100