MiroThinker-1.7 & H1: Towards Heavy-Duty Research Agents via Verification
MiroMind Team, S. Bai, L. Bing, L. Lei, R. Li, X. Li, X. Lin, E. Min, L. Su, B. Wang, L. Wang, L. Wang, S. Wang, X. Wang, Y. Zhang, Z. Zhang, G. Chen, L. Chen, Z. Cheng, Y. Deng, Z. Huang, D. Ng
2026-03-18
Summary
This paper introduces two new AI agents, MiroThinker-1.7 and MiroThinker-H1, designed to tackle complicated problems that require a lot of thinking and planning, like doing in-depth research or analyzing complex data.
What's the problem?
Current AI agents often struggle with tasks that need multiple steps of reasoning and can make mistakes along the way, especially when dealing with information from the internet or specialized fields like science and finance. They aren't always reliable when they need to think through a problem over a long period of time and can sometimes jump to conclusions without solid evidence.
What's the solution?
The researchers built MiroThinker-1.7 to be more careful and organized in each step of its thinking process. It focuses on planning, understanding the context, and using tools effectively. Then, they created MiroThinker-H1, which goes even further by checking its own work as it goes, both at each individual step and by looking at the overall reasoning process to make sure the final answer makes sense and is supported by evidence. They also made the first version, MiroThinker-1.7, and a smaller version available for others to use.
Why it matters?
These new agents represent a significant step forward in AI's ability to perform complex research and analysis. They achieve better results than previous AI systems on challenging tasks and provide a more reliable way to get answers that are backed up by evidence, which is crucial for fields like science, finance, and any situation where accuracy is important.
Abstract
We present MiroThinker-1.7, a new research agent designed for complex long-horizon reasoning tasks. Building on this foundation, we further introduce MiroThinker-H1, which extends the agent with heavy-duty reasoning capabilities for more reliable multi-step problem solving. In particular, MiroThinker-1.7 improves the reliability of each interaction step through an agentic mid-training stage that emphasizes structured planning, contextual reasoning, and tool interaction. This enables more effective multi-step interaction and sustained reasoning across complex tasks. MiroThinker-H1 further incorporates verification directly into the reasoning process at both local and global levels. Intermediate reasoning decisions can be evaluated and refined during inference, while the overall reasoning trajectory is audited to ensure that final answers are supported by coherent chains of evidence. Across benchmarks covering open-web research, scientific reasoning, and financial analysis, MiroThinker-H1 achieves state-of-the-art performance on deep research tasks while maintaining strong results on specialized domains. We also release MiroThinker-1.7 and MiroThinker-1.7-mini as open-source models, providing competitive research-agent capabilities with significantly improved efficiency.