
system1-agents
Prebuilt and custom agents that run on small, fast decision models
- Pricing
- Open source
- Platforms
- CLI
- GitHub stars
- 184
- Last commit
- 1 day ago
- Language
- Python
- License
- Apache-2.0
About system1-agents
system1-agents is an open-source agent for multi-agent frameworks. It is available on CLI.
The idea here is to replace a chat model in the decision loop with a smaller System 1 model, and the README backs this with its own numbers. In its table, Jev took 35.7 s against 138.7 s for a chat model on a WebVoyager Allrecipes task, and cost $0.091 against $1.49; other rows cover a ticket router, Windows Calculator, ALFWorld and several games.
You describe a task to Claude Code or Codex, which either runs a prebuilt agent or builds a new one on Jev, Laya or Cua-S1 Nano. Prebuilt agents span browser use, computer use, robotics and games. The README also mentions OmniJev, a model that reads the screen.
Code is in Python under Apache-2.0, with replays and a benchmarks page in the repository.
What it does
- Prebuilt agents
- Ready-to-run agents for browser, computer use, robotics and games.
- Agent builder
- Claude Code or Codex can generate a new agent for a task.
- Three decision models
- Works with Jev, Laya and Cua-S1 Nano.
- Published benchmarks
- Speed and cost tables compared with a chat model.
- Screen-reading model
- OmniJev reads the screen for computer-use tasks.
Good for
- Speed up a repetitive browser task
- Route support tickets into queues
- Automate a desktop app through its screen
- Prototype a game-playing agent
Pricing
Open source
Apache-2.0
system1-agents FAQ
How much faster is it?
The README reports up to 6x faster and 25x cheaper than a chat model at the same score, based on its own runs.
How do I create an agent?
Describe the task to Claude Code or Codex.
Which languages and license?
Python, Apache-2.0.
Is system1-agents free?
Yes. system1-agents is open source, so you can use and self-host it for free.
What are the best alternatives to system1-agents?
The closest alternatives to system1-agents are hermes-agent, deepseek-harness, langchain, deer-flow and ruflo.
Alternatives to system1-agents

hermes-agent
AgentOpen-source self-improving agent with memory, scheduling and many chat gateways
- Freemium
- Open-source agent
- 264K GitHub stars
- AI models and LLM tooling

deepseek-harness
AgentOpen-source agent harness from DeepSeek, built on a plugin architecture
- Open source
- Open-source agent
- 282K GitHub stars
- Multi-agent frameworks

langchain
AgentPython library for building agents from a model, tools, prompt and middleware
- Open source
- Agent framework
- 150K GitHub stars
- AI models and LLM tooling

deer-flow
AgentOpen-source super agent harness that researches, codes and creates with sandboxes
- Open source
- Open-source agent
- 85.4K GitHub stars
- AI models and LLM tooling

ruflo
AgentAgent meta-harness that adds swarms, memory and MCP tools to Claude Code and Codex
- Open source
- Agent framework
- 81.9K GitHub stars
- Knowledge and memory

autogen
AgentArchived Microsoft framework for multi-agent AI applications in Python
- Open source
- Agent framework
- 61.9K GitHub stars
- Multi-agent frameworks