Skip to main content

​
AI models

AI model settings control where Finseo runs your tracking prompts. More models give broader coverage, but they also increase the number of AI answers generated per prompt.

​
Available models

Finseo supports the following model channels in the dashboard:
Model IDDescription
chatgptChatGPT GUI/UI-style tracking.
chatgpt_guiChatGPT API-style tracking.
perplexityPerplexity answers and source citations.
ai_overviewGoogle AI Overview-style result tracking.
google_ai_modeGoogle AI Mode-style result tracking where available.
claudeClaude answers.
grokGrok answers.
mistralMistral answers.
deepseekDeepSeek answers.
geminiGemini answers.

​
Default model selection

New accounts start with these defaults: 
{
  "chatgpt": true,
  "perplexity": true,
  "ai_overview": true,
  "chatgpt_gui": false,
  "google_ai_mode": false,
  "claude": false,
  "grok": false,
  "mistral": false,
  "deepseek": false,
  "gemini": false
}
This gives you coverage across ChatGPT, Perplexity, and Google-style AI search without running every possible model.

​
Account defaults vs project settings

Finseo supports two levels of model settings:
  • Account defaults apply when a project has no custom model settings.
  • Project-specific settings override the account defaults for one project.
Use project-specific settings when different brands, markets, or clients need different model coverage.

​
Change enabled models

  1. Open Models.
  2. Select the project if you want project-specific settings.
  3. Toggle the models you want to track.
  4. Save the settings.
When you save project-specific settings, Finseo syncs existing tracking prompts in that project to the enabled model set. New prompts use the same enabled models by default.

​
Model limits

Most plans can enable up to 3 models per prompt. Agency plans and custom enterprise quotes can allow more models. If you hit the model limit, disable one model before enabling another. The limit is enforced when saving model preferences and when adding new prompts.
For agency accounts, model count affects projected monthly AI answers. Adding prompts across many models increases usage because each prompt is run once per enabled model.

​
How model count affects results

The number of generated answers is roughly: 
tracked prompts × enabled models × scheduled runs
Example: 
100 prompts × 3 models = 300 AI answers per run
100 prompts × 6 models = 600 AI answers per run
Use more models when you need broad market coverage. Use fewer models when you want focused tracking, faster analysis, or lower usage.

​
Choosing a model mix

Start with:
  • chatgpt for broad assistant visibility.
  • perplexity for citation-heavy AI search behavior.
  • ai_overview for Google AI result visibility.
Then add more models when they matter to your audience:
  • Add claude for B2B, enterprise, technical, or research-heavy markets.
  • Add gemini or google_ai_mode when Google AI surfaces are strategically important.
  • Add grok, mistral, or deepseek when your audience uses those platforms.

​
API behavior

When you create prompts through the API, you can pass an explicit models array. 
{
  "prompt": "Best AI visibility tracking tools for agencies",
  "models": ["chatgpt", "perplexity", "ai_overview"],
  "language": "us",
  "tags": ["agency", "comparison"]
}
If you omit models, Finseo uses the enabled project settings. If no project settings exist, it uses the account defaults. See the Create Prompt API for the full reference.