Provider coverage
14 providers, queried through their APIs — never by scraping a consumer chat window. API samples can differ from what one person sees in the consumer app. Every stored answer keeps the exact model id.
How plans pick providers
- Free always runs ChatGPT with web search. One provider, 5 questions.
- Plus runs all 5 of its catalog on every audit: ChatGPT, Claude, Gemini, Perplexity, and Mistral.
- Pro can pick any 10 of 14 for one audit. Swap in the picker. A missing API key on our side marks that provider partial instead of dropping it quietly.
- Growth (waitlist) would run the 8 most-used consumer AIs. It is not on sale.
Coverage matrix
| Provider | How we query | Search / grounding | Free | Plus | Pro |
|---|---|---|---|---|---|
| ChatGPT | OpenAI Responses API | Web search tool on | Yes | Yes | Yes |
| Claude | Claude via Amazon Bedrock API | API only — citations labelled as model-suggested unless independently grounded | — | Yes | Yes |
| Gemini | Gemini generateContent API | Official API; grounding URLs stored when returned | — | — | Yes |
| Perplexity | Perplexity API | Provider-grounded answers | — | — | Yes |
| Grok | xAI API | API only | — | Yes | Yes |
| DeepSeek | DeepSeek API | API only | — | — | Yes |
| Mistral | Mistral via Amazon Bedrock API | API only | — | Yes | Yes |
| Kimi | Kimi API | API only | — | — | Yes |
| Nova | Amazon Nova via Bedrock API | API only | — | — | Yes |
| Groq | Groq API | API only | — | — | Yes |
| MiniMax | MiniMax API | API only | — | — | Yes |
| Sarvam | Sarvam API | API only | — | — | Yes |
| Qwen | Qwen API | API only | — | — | Yes |
| Llama Search | Official API | API only | — | Yes | Yes |
Buyers see ChatGPT once. The free and paid audits use the ChatGPT route that has web search on.
Update log
This matrix matches the catalog shipped with methodology v1.2.0. We do not have public dates for when each provider was first added. When a provider is added or a plan mapping changes, this page and the methodology page update together.
- Current catalog: ChatGPT, Claude, Gemini, Perplexity, Grok, DeepSeek, Mistral, Kimi, Nova, Groq, MiniMax, Sarvam, Qwen, Llama Search.
- Scoring, prompt rules, and limitations stay on Methodology.