# General Reasoning

Long-horizon reinforcement learning and open agent research.

- Website: https://gr.inc/
- Category: RL startup
- Type: Research lab
- Capabilities: Long-horizon, Simulation, Open source
- Reviewed: 2026-09-09

## Overview

General Reasoning researches agents that learn to solve extended tasks. Its projects include OpenReward, BackSearch, and KellyBench, covering reward signals, reproducible search, and agent evaluation.

## Products & services

- Long-horizon and multi-agent learning research.
- Open tools and benchmarks for agent development.

## Sources & references

- [General Reasoning products and research](https://gr.inc/)