The Year in AI Papers
The year/Independent research

Paper 2511.11793

MiroThinker: Pushing the Performance Boundaries of Open-Source Research Agents via Model, Context, and Interactive Scaling

MiroMind Team, Song Bai, Lidong Bing, Carson Chen, Guanzheng Chen, Yuntao Chen, Zhe Chen, Ziyi Chen, Jifeng Dai, Xuan Dong, Wenhan Dou, Yue Deng, et al.

Published
Nov 2025
Research lab
Independent
Citations
62
GitHub
8.4K stars

01 In brief

Summary

MiroThinker v1.0 is an open-source research agent that introduces interaction scaling as a third performance dimension, alongside model size and context length.

It trains models via reinforcement learning to handle deeper and more frequent agent–environment interactions, using a 256K context window and up to 600 tool calls per task.

The 72B variant achieves state-of-the-art results among open-source agents: 81.9% on GAIA, 37.7% on HLE, 47.1% on BrowseComp, and 55.6% on BrowseComp-ZH, approaching commercial systems like GPT-5-high.

The model uses a ReAct paradigm with tools for execution, file management, and retrieval, plus recency-based context retention.

Training involves SFT, DPO, and GRPO, with data from multi-document QA and agentic trajectory synthesis.

Interactive scaling consistently improves performance as interaction depth increases.

Limitations include tool-use inefficiency, overlong reasoning chains, language mixing, and limited sandbox capability.

Variants at 8B, 30B, and 72B are released with code and weights.

02 From the paper

Abstract

We present MiroThinker v1.0, an open-source research agent designed to advance tool-augmented reasoning and information-seeking capabilities. Unlike previous agents that only scale up model size or context length, MiroThinker explores interaction scaling at the model level, systematically training the model to handle deeper and more frequent agent-environment interactions as a third dimension of performance improvement. Unlike LLM test-time scaling, which operates in isolation and risks degradation with longer reasoning chains, interactive scaling leverages environment feedback and external information acquisition to correct errors and refine trajectories. Through reinforcement learning, the model achieves efficient interaction scaling: with a 256K context window, it can perform up to 600 tool calls per task, enabling sustained multi-turn reasoning and complex real-world research workflows. Across four representative benchmarks-GAIA, HLE, BrowseComp, and BrowseComp-ZH-the 72B variant achieves up to 81.9%, 37.7%, 47.1%, and 55.6% accuracy respectively, surpassing previous open-source agents and approaching commercial counterparts such as GPT-5-high. Our analysis reveals that MiroThinker benefits from interactive scaling consistently: research performance improves predictably as the model engages in deeper and more frequent agent-environment interactions, demonstrating that interaction depth exhibits scaling behaviors analogous to model size and context length. These findings establish interaction scaling as a third critical dimension for building next-generation open research agents, complementing model capacity and context windows.