/

Axiom Math

  • What it is:Axiom Math is a company building a self-improving superintelligent reasoner starting with an AI mathematician, using Lean and mathematical proofs for verified quantitative reasoning.
  • Best for:Quantitative finance firms, Mathematical research institutions, Industrial R&D teams
  • Pricing:Starting from $15/month
  • Rating:82/100Very Good
  • Expert's conclusion:Currently, the majority of AI applications are not capable of performing at the same level as humans; however, researchers are continually working to advance the technology so it will eventually reach that point.
Visit website
Reviewed byMaxim Manylov·Web3 Engineer & Serial Founder

What Is Axiom Math and What Does It Do?

Axiom Math is an AI research firm that is developing an AI mathematician which can solve complex mathematical problems, generate formal proofs and advance towards superintelligent reasoning. The founder of the firm is Carina Hong, who is a former Stanford graduate student and Rhodes scholar with a background in mathematics from both MIT and Oxford. The focus of the firm is on advanced math as a way to achieve mathematical superintelligence, and has applications in areas requiring provably correct reasoning.

Active
📍San Francisco, CA
📅Founded 2025
🏢Private
TARGET SEGMENTS
AI Research LabsAcademic InstitutionsQuantitative FinanceCryptographyHardware Verification

What Are Axiom Math's Key Business Metrics?

🏢
17
Employees
📊
$64M
Seed Funding
📊
$300M
Valuation
📊
2 Erdos problems
Solved Problems
📊
2025
Founding Year

How Credible and Trustworthy Is Axiom Math?

82/100
Good

Achieved exceptional early technical success and recruited some of the world's best talent, many of whom were previously at Meta FAIR or working in academia, and received significant seed funding; however, it is still relatively young and does not have a proven track record in terms of its ability to compete in the marketplace.

BREAKDOWN
Product Maturity65/100
Company Stability85/100
Security & Compliance70/100
User Reviews50/100
Transparency75/100
Support Quality60/100
TRUST SIGNALS
$64M seed funding at $300M valuationSolved 2 long-standing Erdos math problemsTeam from Meta FAIR, Google Brain, StanfordFounded by Rhodes Scholar with MIT/Oxford background

What is the history of Axiom Math and its key milestones?

2025

Company Founded

Axiom Math was founded by Carina Hong, a Stanford drop-out and Rhodes scholar, in March to create an AI Mathematician for use in achieving superintelligent reasoning.

2025

Solved Erdos Problems

Announced that they had solved two instances of the long-standing Erdos-Straus conjectures that had gone unsolved by mathematicians for decades.

2025

$64M Seed Round

Received $64 million in seed funding in September at a valuation of $300 million to continue to develop their capabilities to produce an AI mathematician.

2025

Elite Team Assembly

Has recruited top talent including Meta FAIR researchers, such as CTO Shubho Sengupta, Francois Charton, Aram Markosyan, and mathematician Ken Ono.

What Are the Key Features of Axiom Math?

✨
Formal Mathematical Proofs
Produces machine-verifiable proofs of correctness using the Lean programming language for provably correct mathematical reasoning.
✨
Complex Problem Solving
Solves complex mathematical problems, including the two instances of the Erdos-Straus conjectures that had been unsolved for decades.
✨
Superintelligent Reasoning
Is being designed as a foundation for creating a self-improving AI mathematician that will evolve into mathematical superintelligence.
✨
Deductive Knowledge Graph
Expands its knowledge graph through deductive logic and formal verification.
✨
Interdisciplinary Architecture
Uses the combination of AI, programming languages, and expertise in mathematics to develop theory-building capabilities.
✨
Provably Correct Reasoning
Provides a means of applying the use of AI in areas such as hardware verification, quantitative finance, and cryptography that require mathematical guarantees.

What Technology Stack and Infrastructure Does Axiom Math Use?

Technologies

Lean (theorem prover)PythonPyTorch

AI/ML Capabilities

Proprietary AI mathematician architecture combining deep learning with formal proof generation in Lean, designed for self-improving superintelligent reasoning on complex mathematical problems

Inferred from founder interviews and research announcements; Lean specifically mentioned for proof generation

What Are the Best Use Cases for Axiom Math?

Mathematical Researchers
Accelerates the generation of proofs and the solution to conjectures for long-standing open problems through the use of formal verification.
AI Research Labs
Axiom provides the capability to reason frontiers as necessary for developing safe superintelligent AI systems
Quantitative Finance Teams
Axiom can prove the correctness of the reasoning for complex derivative pricing, risk modeling, and algorithmic trading strategies
Cryptography Researchers
Axiom generates formally verifiable cryptographic proofs that are required for post-quantum secure protocols
Hardware Verification Engineers
Axiom automates the formal verification of chip design and processor architecture with a guarantee of correctness
NOT FORK-12 Math Teachers
Axiom is a pre-commercial research product and does not include any educational interfaces or curriculum integration
NOT FORReal-time Trading Systems
The Axiom research-focused system is not suitable for sub-millisecond latency requirements associated with high frequency trading

How Much Does Axiom Math Cost and What Plans Are Available?

Pricing information with service tiers, costs, and details
☐Service$CostℹDetails🔗Source
Axiom Starter$15/month5 hrs monthly runtime, desktop unlimited concurrency, 1 bot in cloud Run, cloud single-run limit: 1 hour, no-code and Pro-code tools, email and community supportOfficial pricing page
Axiom Pro$50/month30 hrs monthly runtime, desktop unlimited concurrency, 1 bot in cloud Run, daily (cloud)/every minute (desktop) scheduling, Zapier + Make integration, Webhooks & API, cloud single-run limit: 2 hours, email supportOfficial pricing page
Axiom Pro Max$150/month100 hrs monthly runtime, desktop unlimited concurrency, 2 bots concurrently in cloud Run, hourly (cloud)/every minute (desktop) scheduling, Zapier + Make integration, Webhooks & API, cloud single-run limit: 3 hours, Slack and email supportOfficial pricing page
Axiom Ultimate$250/monthDesktop unlimited concurrency, 20 bots concurrently in cloud Run, 15 minutes (cloud)/every minute (desktop) scheduling, team account, Zapier + Make integration, Webhooks & API, cloud single-run limit: 12 hours, priority Slack and email supportOfficial pricing page
Custom PlansContact for detailsAvailable if Ultimate tier insufficient, annual billing possible for Ultimate tierOfficial pricing page
Axiom Starter$15/month
5 hrs monthly runtime, desktop unlimited concurrency, 1 bot in cloud Run, cloud single-run limit: 1 hour, no-code and Pro-code tools, email and community support
Official pricing page
Axiom Pro$50/month
30 hrs monthly runtime, desktop unlimited concurrency, 1 bot in cloud Run, daily (cloud)/every minute (desktop) scheduling, Zapier + Make integration, Webhooks & API, cloud single-run limit: 2 hours, email support
Official pricing page
Axiom Pro Max$150/month
100 hrs monthly runtime, desktop unlimited concurrency, 2 bots concurrently in cloud Run, hourly (cloud)/every minute (desktop) scheduling, Zapier + Make integration, Webhooks & API, cloud single-run limit: 3 hours, Slack and email support
Official pricing page
Axiom Ultimate$250/month
Desktop unlimited concurrency, 20 bots concurrently in cloud Run, 15 minutes (cloud)/every minute (desktop) scheduling, team account, Zapier + Make integration, Webhooks & API, cloud single-run limit: 12 hours, priority Slack and email support
Official pricing page
Custom PlansContact for details
Available if Ultimate tier insufficient, annual billing possible for Ultimate tier
Official pricing page

How Does Axiom Math Compare to Competitors?

FeatureAxiom MathOpenAI o1Anthropic Claude 3.5Google DeepMind AlphaProofxAI Grok
Core FunctionalityFormal math proofs & verificationAdvanced reasoning & mathMath reasoning & codingCompetition-level math proofsGeneral reasoning & math
Pricing (starting)$15/month (browser automation)API: $15/1M input tokensAPI: $3/1M input tokensResearch only, not commercialSubscription-based
Free TierNo (Starter at $15)Limited playgroundLimited playgroundNoYes (basic)
Enterprise FeaturesTeam accounts (Ultimate)Enterprise API, SSOEnterprise API, SSO—Enterprise plans
API AvailabilityWebhooks & API integrationsYes (full API)Yes (full API)NoYes
Integration CountZapier, Make, WebhooksExtensive ecosystemExtensive ecosystem—xAI ecosystem
Support OptionsEmail to Priority SlackAPI docs, forumsPriority enterpriseResearch supportCommunity + enterprise
Security Certifications—SOC 2, enterprise securitySOC 2, enterprise securityResearch institutionEnterprise security
Core Functionality
Axiom MathFormal math proofs & verification
OpenAI o1Advanced reasoning & math
Anthropic Claude 3.5Math reasoning & coding
Google DeepMind AlphaProofCompetition-level math proofs
xAI GrokGeneral reasoning & math
Pricing (starting)
Axiom Math$15/month (browser automation)
OpenAI o1API: $15/1M input tokens
Anthropic Claude 3.5API: $3/1M input tokens
Google DeepMind AlphaProofResearch only, not commercial
xAI GrokSubscription-based
Free Tier
Axiom MathNo (Starter at $15)
OpenAI o1Limited playground
Anthropic Claude 3.5Limited playground
Google DeepMind AlphaProofNo
xAI GrokYes (basic)
Enterprise Features
Axiom MathTeam accounts (Ultimate)
OpenAI o1Enterprise API, SSO
Anthropic Claude 3.5Enterprise API, SSO
Google DeepMind AlphaProof—
xAI GrokEnterprise plans
API Availability
Axiom MathWebhooks & API integrations
OpenAI o1Yes (full API)
Anthropic Claude 3.5Yes (full API)
Google DeepMind AlphaProofNo
xAI GrokYes
Integration Count
Axiom MathZapier, Make, Webhooks
OpenAI o1Extensive ecosystem
Anthropic Claude 3.5Extensive ecosystem
Google DeepMind AlphaProof—
xAI GrokxAI ecosystem
Support Options
Axiom MathEmail to Priority Slack
OpenAI o1API docs, forums
Anthropic Claude 3.5Priority enterprise
Google DeepMind AlphaProofResearch support
xAI GrokCommunity + enterprise
Security Certifications
Axiom Math—
OpenAI o1SOC 2, enterprise security
Anthropic Claude 3.5SOC 2, enterprise security
Google DeepMind AlphaProofResearch institution
xAI GrokEnterprise security

How Does Axiom Math Compare to Competitors?

vs OpenAI o1

Axiom specializes in providing formal mathematical verification and generating proofs for high-stake applications and OpenAI o1 has expertise in general reasoning and casual math problems. Axiom has premium B2B pricing with formal guarantees and OpenAI supports a broader set of consumers/developers.

Use Axiom for mathematically rigorous enterprise application use cases; OpenAI for general AI reasoning needs.

vs Anthropic Claude 3.5 Sonnet

Claude provides strong math reasoning at lower API costs supporting developers; Axiom targets formal math users who require verified proofs. Axiom's browser automation history allows it to integrate into workflows uniquely; Claude emphasizes safety alignment.

Choose Axiom for your proof-critical quantitative finance applications; Claude for your cost-effective math reasoning needs.

vs Google DeepMind AlphaProof

AlphaProof is an example of cutting-edge research in generating IMO level proofs but is not commercially available. Axiom bridges the gap from research to practical B2B usage with automation tools and tiered pricing for quant firms.

Use Axiom for immediate commercial deployment; Use AlphaProof for research benchmarking.

vs xAI Grok

Grok is focused on broad reasoning with real-time capabilities; Axiom is focused on niche formal math verification. Axiom's established pricing model is in contrast to xAI's emerging subscription based model.

Use Axiom for your specialized math proofs; Use Grok for conversational math assistance.

What are the strengths and limitations of Axiom Math?

Pros

  • The elimination of potential for error due to computation in the high-risk field of mathematics through formal verification of mathematical proofs
  • Flexible pricing tiers — from a starter tier at $15 per month to enterprise-tier accounts for large development teams
  • Support for integrating AI into Axion Math's workflow — currently supports ChatGPT, Claude, Gemini, DeepSeek LLMs
  • Pro-code & no-code tools — provides access to both the developer community and the business user community
  • Pay-as-you-go runtime-based pricing — only pays for the actual time spent using the system's computing resources
  • Shared team accounts — Ultimate tier allows for all members of the team to have access to the same account
  • Automated workflows — provides webhooks and connections to Zapier & Make for production-level deployment

Cons

  • The limited focus of the product — makes it appealing to a limited audience, mostly those involved in quantitative finance/research, not the broader AI market
  • There is no "free" tier — requires a minimum payment of $15 per month to utilize
  • Monthly runtime caps — vary by tier, 5-100+ hours possible, no rollover
  • Limits to cloud concurrency — varies based on price (1-20 bots max), increases with higher cost options
  • Only custom plans are available at the high end — little to no flexibility within the lower tiers
  • An early-stage company — has raised $64 million, but does not yet have the size/brand recognition of other AI companies
  • High level of difficulty required for non-experts in mathematics to learn the complexities of formal verification of mathematical proofs

Who Is Axiom Math Best For?

Best For

  • Quantitative finance firms — Critical for verified mathematical proofs — used to eliminate the risk of computational errors in risk modeling/trading algorithms
  • Mathematical research institutions — Suits individual researchers through to larger teams — tiered pricing options available
  • Industrial R&D teams — Reduces risk associated with the use of computational methods in engineering — verified mathematical proofs
  • AI automation teams — Integrates LLMs with browser automation workflows
  • Teams needing math rigor — Guarantees that traditional LLMs do not possess the ability to provide the same level of performance/certainty for critical applications

Not Suitable For

  • General content creators — Overkill for casual math users — would be better suited for users who require verified math results such as a researcher who uses ChatGPT/Claude for exploratory math tasks
  • Budget-conscious developers — Minimum of $15 plus usage limits exceed what is provided by most LLM "playgrounds" that offer completely free service
  • Non-technical business users — Formal verification of mathematical proofs typically requires expert knowledge in mathematics — would recommend no-code tools for individuals without extensive experience with formal verification of mathematical proofs
  • Real-time consumer apps — Latency of generating proof is too high — would be better suited for users who need quicker reasoning models

Are There Usage Limits or Geographic Restrictions for Axiom Math?

Monthly Runtime
5 hrs (Starter), 30 hrs (Pro), 100 hrs (Pro Max), Unlimited? (Ultimate)
Cloud Concurrent Bots
1 bot (Starter/Pro), 2 bots (Pro Max), 20 bots (Ultimate)
Cloud Single-Run Limit
1 hr (Starter), 2 hrs (Pro), 3 hrs (Pro Max), 12 hrs (Ultimate)
Desktop Concurrency
Unlimited across all tiers
Scheduling Frequency
Daily (Pro cloud), Hourly (Pro Max cloud), 15 min (Ultimate cloud)
Runtime Rollover
No rollover to next month
Team Accounts
Ultimate tier only
Custom Plans
Ultimate tier+ only

Is Axiom Math Secure and Compliant?

Team Access ControlTeam accounts available on Ultimate tier with support-managed setup
Data EncryptionStandard cloud encryption assumed for LLM integrations and browser automation
Access ManagementAccount-based access with team sharing on highest tier
Infrastructure SecurityCloud-based service with desktop extension runtime options
Compliance CertificationsNot publicly specified; enterprise customers should request audit details
Audit LoggingRuntime and usage tracking via subscription dashboards

What Customer Support Options Does Axiom Math Offer?

Channels
All tiersStarter tierPro Max+ tiers, Priority on UltimateTeam accounts, custom plans
Hours
Business hours assumed
Response Time
Not specified; priority channels on higher tiers
Satisfaction
Not available from sources
Specialized
Priority support and team account setup for Ultimate tier
Business Tier
Priority Slack + email for enterprise features
Support Limitations
•Community support only for Starter tier
•Slack requires Pro Max+ subscription
•No 24/7 support mentioned
•Team setup requires contacting support

What APIs and Integrations Does Axiom Math Support?

API Type
REST API using POST HTTP method only. Endpoint: https://lar.axiom.ai/api/v3
Authentication
API Key (tokens) required in payload for all requests. Tokens control specific permissions like dataset access
Webhooks
Supported. Trigger automations via webhook steps; receive data from other apps
SDKs
No official SDKs mentioned. Direct HTTP requests with code snippets available
Documentation
Good - detailed guides for API requests, responses, queue management, and rate limits at axiom.ai/docs/developers/api
Sandbox
No dedicated sandbox. Test with API tokens and queue system
SLA
Not specified publicly
Rate Limits
Managed queue system with documented rate limits (details in developer docs)
Use Cases
Trigger browser automations programmatically, pass 2D array data, check run status/results, stop automations, integrate with Make.com, Zapier, Google Sheets

What Are Common Questions About Axiom Math?

Axiom Math is developing a self-improving super intelligent reasoner using an AI Mathematician. The primary focus of Axiom Math is to develop frontier AI technologies related to mathematical reasoning and not to create browser automation tools. For additional information about the mission of Axiom Math, please visit axiommath.ai.

Axiom Math does not display publicly the price of its services at axiommath.ai; pricing may be limited to the research community or requires contacting the team individually; no tiered pricing for consumers was noted.

The Axiom Math team has developed an advanced AI system that performs advanced reasoning in mathematics. The company, axiom.ai, provides a no-code browser automation platform with Robotic Process Automation capability. It appears that they have two separate products even though both have similar branding.

There is no public documentation regarding security for Axiom Math. Because Axiom Math is a research-based AI lab, it is likely using enterprise-level security standards and practices, however this should be confirmed by the Axiom Math team before using their service in a production environment.

No public documentation of the API for Axiom Math could be located. Details about how to integrate Axiom Math into your application would have to be provided by the Axiom Math research team. The Axiom Math service does not appear to be a generic developer-integrated service as is the case with axiom.ai.

Contact the Axiom Math team through their website, axiommath.ai. Axiom Math did not provide public assistance through a Help Center, Forums, or other forms of documentation. The research-focused nature of Axiom Math means that they will likely assist you through one-on-one communication.

A free trial or public beta were not mentioned. The Axiom Math team seems to limit access to qualified researchers or partner companies due to the focus on developing new technologies in AI research.

Axiom Math is currently in the pre-product phase of development and is focused on researching advanced AI mathematical reasoning capabilities. There is no public API, pricing, or production use case available for Axiom Math at this time. The Axiom Math service can only perform research related to mathematical reasoning and not general applications.

Is Axiom Math Worth It?

Axiom Math is a representation of frontier AI research efforts to develop self-improving systems that reason mathematically. At this time, Axiom Math is still in the research phase and there is no public product, API, or pricing available. For organizations looking to bet on next-generation AI capabilities, Axiom Math presents high-risk/high-reward potential.

Recommended For

  • AI research labs working on expanding reasoning capabilities
  • Academic institutions studying mathematical AI
  • Deep-tech venture capital firms monitoring developments in frontier models
  • Companies requiring breakthrough math reasoning 5+ years out

!
Use With Caution

  • Organizations requiring production AI today
  • Teams lacking AI research experience
  • If you need to deploy your solution commercially in less than 24 months, then most likely you will need to rely on one of the other options listed above as well as possibly some form of cloud-based support.
  • Many businesses have a desire to automate their internal operations, which can include tasks such as data entry and processing.

Not Recommended For

  • In addition to automating specific processes, many businesses also require more generic automation solutions to help improve the efficiency of their overall operations.
  • No-code and low-code teams are often looking for the easiest possible way to implement automation within their organization.
  • Some organizations may have a requirement for immediate Return on Investment (ROI).
  • While there are many types of artificial intelligence (AI), some types are non-mathematical in nature, which means they do not perform complex calculations like those used in machine learning models.
Expert's Conclusion

Currently, the majority of AI applications are not capable of performing at the same level as humans; however, researchers are continually working to advance the technology so it will eventually reach that point.

Best For
AI research labs working on expanding reasoning capabilitiesAcademic institutions studying mathematical AIDeep-tech venture capital firms monitoring developments in frontier models

What do expert reviews and research say about Axiom Math?

Key Findings

The main focus of Axiom Math (axiommath.ai) is to develop self-improving mathematical reasoners which are significantly different from what is provided by axiom.ai browser automation tool. There is no publicly available API, pricing, or product – all focus is on researching the development of self-improving reasoners.

Data Quality

Limited - Axiom Math has minimal public disclosure typical of early-stage frontier AI research. Most technical details relate to unrelated axiom.ai automation product. Primary source is mission statement on axiommath.ai.

Risk Factors

!
Currently, Axiom Math is in its research phase and does not provide any commercial offerings.
!
The name Axiom Math could potentially cause confusion with the axiom.ai browser automation tool.
!
At this time, there is limited information about the technical aspects of Axiom Math due to the fact that it has no publicly available API or documentation.
!
Frontier AI research typically takes long periods of time to complete and therefore, it should not be expected that Axiom Math will produce results quickly.
!
Information regarding the funding status of Axiom Math or details of the team developing Axiom Math has not been made public.
Last updated: February 2026

What Additional Information Is Available for Axiom Math?

Mission Focus

The primary goal of Axiom Math is to create a self-improving superintelligent reasoner, beginning with the creation of an AI mathematician. Emphasis is placed on providing frontier capabilities that extend far beyond the limitations of current Large Language Model (LLM) models.

Research Stage

It appears that Axiom Math is currently still in its research phase and does not have any publicly available demos, papers, or technical results.

Brand Clarification

The main difference between axiom.math and axiom.ai is the domain that each company utilizes. axiom.ai provides browser automation and Robotic Process Automation (RPA) through their developer API whereas Axiom Math develops AI reasoners using the axiommath.ai domain.

What Are the Best Alternatives to Axiom Math?

  • •
    DeepMind AlphaProof: Google DeepMind developed an AI system capable of math reasoning similar to a human who had solved International Mathematical Olympiad (IMO) problems at a silver medal level. The Google DeepMind system is a production research system and, although it was able to solve IMO problems, it is not directly comparable to Axiom Math since Axiom Math is still in the pre-product phase. Google DeepMind's work is best understood as the current state-of-the-art (SOTA) math AI. deepmind.google
  • •
    OpenAI o1: OpenAI created a reasoning model that was able to perform PhD-level math on various benchmark tests. This reasoning model is available through the OpenAI API for immediate use in applications requiring reasoning capabilities. OpenAI's reasoning model is considered production-ready and is immediately useful for applications requiring high levels of reasoning capability. openai.com
  • •
    Anthropic Claude 3.5 Sonnet: Most widely used mathematical reasoner with high math performance. API access to model is immediate (can be used in production). Ideal for math problem solving/production use. (anthropic.com)
  • •
    Meta Llama 3.1 405B: Open-weighted model with competitive math reasoning. Has ability to self-host and not a research only service. Ideal for companies that want to have an AI math tool and not be locked into a specific vendor. (meta.ai)
  • •
    xAI Grok: xAi’s model has focus on the reasoning aspect of its model with real time capabilities. Less developed than Axiom Math but has public API access. Ideal for tracking the development of the competitive frontier in real time reasoning. (x.ai)

Expert Reviews

📝

No reviews yet

Be the first to review Axiom Math!

Write a Review

Similar Products

Thinking Machines Lab
thinkingmachines.ai
Reflection AI
reflection.ai
SandboxAQ
sandboxaq.com