AI Audit
AI security analysis for smart contracts
No audits yet
Enter a contract address to run your first audit
Browse the Public Audit Record
Audits published to the Quantum Audit dashboard stay there as dated, verifiable security records. Browsing every published report is free and needs no account.
Browse Published AuditsAbout the AI Smart Contract Audit
Paste any contract address and Quantum Audit runs an automated security analysis of that contract in under a minute, across Ethereum, BNB Chain, Polygon and Solana. The network is detected from the address format, so there is nothing to configure. Each run cross-references five or more independent on-chain data sources and checks contract verification, ownership and mint authority, honeypot indicators, liquidity pool locks, holder concentration, buy and sell taxes and token age — more than twenty vulnerability categories in total. Where a contract's source code is verified, that source is read and analyzed directly rather than matched against bytecode patterns alone.
The report separates two things that are often conflated. LLM-powered analysis surfaces the findings — graded Critical, High, Medium, Low or Informational, each with a description and a recommendation. The risk score from 0 to 100 is then computed by a separate deterministic, fact-weighted function, so the same contract produces the same score on every run no matter who submits it. Every factor that moved the number is recorded in an auditable breakdown. Read how the risk score is calculated, or browse the public record of published token audits to see finished reports before running your own.
Frequently Asked Questions
How do I audit a smart contract address?
Paste the contract address into the field on this page and start the analysis. The network is detected automatically from the address format and can be overridden manually. Typical analysis takes under 60 seconds and returns severity-graded findings, a deterministic risk score, an executive summary and a downloadable PDF report you can question further in an AI chat.
What does the AI smart contract audit check?
Every run checks contract verification status, ownership and mint authority, honeypot indicators, liquidity pool locks, holder concentration, buy and sell taxes and token age — over twenty vulnerability categories, cross-referenced against five or more independent on-chain data sources. Where verified source code is available, it is read and analyzed directly, and the findings cite the actual code constructs they refer to.
Which networks are supported?
Four networks: the EVM-compatible chains Ethereum, BNB Chain and Polygon, plus Solana. The chain is inferred from the address format — EVM addresses and Solana base58 addresses are distinguishable — and you can override the detected chain before starting the analysis.
Is the AI smart contract audit really free?
Yes. Running an audit on the free tier costs nothing and returns the full report — findings, risk score, breakdown and PDF. A Pro subscription adds Deep Research for complex contracts and the ability to publish an audit to the public dashboard. Browsing already-published audits is free and requires no account at all.
How is the risk score calculated?
In two separate stages. LLM-powered analysis reads the contract and surfaces severity-graded findings; a deterministic, fact-weighted function then computes the risk score from 0 to 100 across seven factor groups — contract administration, holder concentration, liquidity, taxes, token age, severity-weighted findings and per-token variance. The score is not an AI verdict: the same contract always produces the same number, and every contributing factor is listed in an auditable breakdown. The full method is published on the methodology page.
What does the Public Audit toggle do?
Public Audit is a Pro setting. With it switched on, the finished audit is published to the public dashboard as a dated, permanent security record with its own page — full findings, risk score and breakdown, readable by anyone without an account. You can also publish an audit after the fact from the report screen. Any Pro user can audit and publish any token, so a record cannot be frozen in a favorable state. The score is computed the same way regardless of who ran the audit, which is exactly why the record is worth showing.
Does a low risk score mean the contract is safe?
No. A risk score is proof of audit, not proof of safety. It describes the on-chain facts at the moment of analysis and is not investment advice. Scores fall into four bands — 0–20 low, 21–45 medium, 46–70 high, 71–100 critical — and they describe the number and severity of risk indicators found, not a prediction. Contracts get upgraded, liquidity locks expire and holder distributions shift, so re-audit before acting on an older report.
How does auditing a Solana SPL token differ?
SPL tokens are not ERC-20s, so a Solana audit applies a Solana-specific factor set rather than a translated EVM checklist: mint authority, freeze authority, program ownership, LP lock state and Token-2022 extensions such as transfer fees and permanent delegate. Paste an SPL mint address and the Solana factor set is applied automatically.
Can an AI audit replace a manual security review?
They answer different questions. The automated audit covers known vulnerability classes and on-chain risk factors in under a minute, which makes it the right tool for screening a token before buying, or for a first pass during development. Protocols handling significant value also need a manual review, where auditors reason about business logic and economic assumptions no pattern set encodes. An expert manual review is available on the tools page.
