# Free GEO audit

Enter one domain. See whether ChatGPT and DeepSeek name you in AI answers, then unlock weekly tracking when you want the full loop. This audit walks through every answer, not just a score.

## How it works

We ask the assistants buyers use: Your domain is checked against ChatGPT and DeepSeek with two fixed prompts about AI-answer visibility tools.

Results stay shareable for a week: A fresh audit is cached so the same domain can be shared without re-running. High traffic days fall back to the last result or tomorrow's allowance.

Weekly tracking is the product: This page is a snapshot. A Lite trial adds weekly AI-answer visibility with community evidence and gap actions.

## What this checks

- Two fixed prompts asking which generative engine optimization (GEO) tools get recommended.
- Both prompts run against DeepSeek and ChatGPT, four answers total per audit.
- Each answer is scanned for your brand name and domain with a plain word-boundary text match, not a person or a model judging the wording.
- The audit walks through every answer in order, so you see which prompt did or didn't name you, not only a final number.
- A domain's freshest audit is cached for 7 days, so re-running it inside that window returns the same walkthrough instead of spending another check.

## How to read your result

Your score is the share of the four answers that named you, 0 to 100. Below it, each engine's answer is shown in full, so a 25 doesn't just mean "low": it means one of four answers named you, and you can read exactly which one and why. A zero after a real audit means the assistants were asked and didn't say your name, not that something broke. Because the match is a plain text search, an answer that only repeats your brand name from the prompt itself still counts as a mention, so treat this as a starting signal to investigate, not a certified score.

## What to do next

1. Read the exact wording of every answer, not just the score, and note which competitors got named instead of you.
2. Check whether those competitors have content that answers the same question you were asked, then look at what's missing from your own site or docs.
3. A paid plan runs this audit weekly instead of once, alongside developer-community mentions and gap actions.

## Limits

- Checks DeepSeek and ChatGPT only, two of the six engines paid plans track weekly (Claude, Gemini, Grok, and Perplexity are the other four).
- 3 fresh audits per network (IP address) per day; a fresh result is cached for 7 days per domain, so re-checking the same domain shows the same audit.
- On busy days, once 200 fresh audits have run sitewide, new domains get the last cached result or a "try tomorrow" message instead of a fresh run.
- A score of zero means we asked and you were not named, not that the product is broken.

## FAQ

### What is a GEO audit?

A GEO audit checks whether generative assistants name your product when someone asks for tools in your category. MarketHQ's paid plans turn that into weekly tracking.

### Is this really free?

Yes, a limited number of audits per day with no card required. For weekly AI tracking and community monitoring, begin a Lite trial.

### How is this different from MarketHQ?

This page is a one-off snapshot on two assistants. MarketHQ adds more prompts, community evidence, and gap actions for growing brands.

### Why did I get a different score when I ran it again?

Assistants aren't perfectly consistent between runs, and a cached result is served as-is for 7 days rather than re-run every time. Wait for the cache to expire before comparing two runs, and treat a small change as noise rather than a real shift.

## Related

- Free AEO audit: /tools/aeo-audit (same check, AEO framing)
- AI visibility: /use-cases/for-ai-visibility (product use case)
- Best AI visibility tools: /best/ai-visibility-tools-for-developer-tool-companies (commercial shortlist)
- Pricing: /pricing (Lite trial details)