Ax-Prover: A Deep Reasoning Agentic Framework for Theorem Proving in Mathematics and Quantum Physics
Authors: Benjamin Breen, Marco Del Tredici, Jacob McCarran, Javier Aspuru Mijares, Weichen Winston Yin, Kfir Sulimany, Jacob M. Taylor, Frank H. L. Koppens, Dirk Englund
Organizations: Axiomatic AI · Massachusetts Institute of Technology (MIT) · Institut de Ciències Fotòniques (ICFO) · Institució Catalana de Recerca i Estudis Avançats (ICREA)
We present Ax-Prover, a multi-agent system for automated theorem proving in Lean that can solve problems across diverse scientific domains and operate either autonomously or collaboratively with human experts. To achieve this, Ax-Prover approaches scientific problem solving through formal proof generation, a process that demands both creative reasoning and strict syntactic rigor. Ax-Prover meets this challenge by equipping Large Language Models (LLMs), which provide knowledge and reasoning, with Lean tools via the Model Context Protocol (MCP), which ensure formal correctness. To evaluate its performance as an autonomous prover, we benchmark our approach against frontier LLMs and specialized prover models on two public math benchmarks and on two Lean benchmarks we introduce in the fields of abstract algebra and quantum theory. On public datasets, Ax-Prover is competitive with state-of-the-art provers, while it largely outperforms them on the new benchmarks. This shows that, unlike specialized systems that struggle to generalize, our tool-based agentic theorem prover approach offers a generalizable methodology for formal verification across diverse scientific domains. Furthermore, we demonstrate Ax-Prover's assistant capabilities in a practical use case, showing how it enabled an expert mathematician to formalize the proof of a complex cryptography theorem.
Figures & tables
Figure 1: Left: the multi-agent workflow of Ax-Prover. Right: the tool-enhanced reasoning of the Prover.
Figure 2: The main steps performed by the Prover to prove a theorem.
Category
Tools
Project and File Management
lean_build : Compile and build the Lean project lean_file_contents : Get contents of a Lean file lean_declaration_file : Find which file contains a declaration
Diagnostics and Feedback
lean_diagnostic_messages : Compile code and return diagnostic messages lean_goal : Get the current proof goal at a position lean_term_goal : Get goal information for a term lean_hover_info : Get hover information for symbols
lean_leansearch : Search for theorems and lemmas lean_loogle : Search for lemmas by type signature lean_state_search : Search proof states lean_hammer_premise : Use automated theorem proving
Table 1: Lean tools available on lean-lsp-mcp , organized by functionality.
Dataset
Subset
Ax-Prover
Sonnet
DS-Prover
Kimina
NuminaMath-LEAN
solved-K
81%
7%
48%
100% †
solved-H
47%
8%
14%
0% †
unsolved
26%
1%
18%
0% †
total
51%
5%
28%
31%
AbstractAlgebra
easy
72%
10%
26%
12%
intermediate
56%
6%
22%
14%
Table 2: Models’ performance on NuminaMath-LEAN, AA, and QT . † The results on NuminaMath-LEAN for Kimina are reported from [ 47 ] , and were obtained during its RL training phase with, on average, pass@68.
Model
Accuracy
Compute
Hilbert
72% [462]
avg. pass@1840
Seed-Prover
51% [329]
medium
Ax-Prover †
14 % [92]
pass@1 ‡
Goedel-Prover-V2 †
13% [86]
pass@184
DeepSeek-Prover-V2 †
7% [47]
pass@1024
DSP+ †
4% [23]
pass@128
Table 3: Accuracy results on PutnamBench (% and absolute number of solved problems). † denotes open-source [ 58 ] . ‡ Remember that for Ax-Prover, pass@1 is made of multiple steps, see Section 5.1 .
Tool
Calls
edit_file
36.79
lean_diagnostic_messages
30.73
lean_goal
8.17
lean_loogle
5.88
lean_leansearch
4.32
file_contents
3.00
Table 4: Tool usage statistics.
Appendix figures & tables1 asset
Supplementary material from the paper’s appendix.
Appendix
Tactic
Ax-Prover
DeepSeek
Kimina
apply
X
X
X
assumption
X
X
X
by_cases
X
X
X
calc
X
X
X
cases
X
X
X
change
X
Appendix
Table 5: Tactics used by Ax-Prover, DeepSeek, and Kimina. An “X” indicates the model uses the tactic.