Ask AI models
ask_ai_modelsAsks one prompt of the ChatGPT, Claude, Gemini and Perplexity APIs (and Grok, DeepSeek and Muse Spark, only when named; the last two answer from what the model learned, without searching the web) and the ChatGPT and Gemini apps, as a Run on Prompt Explorer does: each answer and the sources it cites, the searches the ChatGPT app ran, and whether each names a brand when one is given. Estimated price: $0.01 to $1.06 an answer from an API model, by model and version, and $0.01 an answer from an app. An answer saved for the same prompt, model, version, web search and country, here or on the page, is read free and not bought again; only answers with nothing saved are bought. Only an app's answer is ever bought again, once out of date or empty, and only with refreshApps; for fresh answers from the API models over time, track the prompt on the Prompt Tracking page, which buys new ones on a schedule. Answers are kept in Prompt Explorer's Saved answers and History; get_saved_prompt_answers reads them without buying. Several answers can take a few minutes; if your client gives up first, the answers that finished are still saved, so read them with get_saved_prompt_answers before calling again and buying them twice. Each call may spend up to the workspace's agent spend limit ($2.00 unless changed in Settings) without asking: it holds at most its estimate plus a margin while it runs and is charged what it bought, never more. A call whose maximum is above the limit is refused before anything is bought, with its estimate and maximum: ask the user, then call again with approvedMaxCostUsd set to the amount they approved.
Tool call template
Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.
{
"id": 1,
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"arguments": {
"projectId": "<projectId>",
"prompt": "<prompt>"
},
"name": "ask_ai_models"
}
}Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| apps | No | The consumer apps to ask as a person uses them; the ChatGPT app also reports the searches it ran. Defaults to none. | |
| brand | No | A brand to check the answers for, free: whether each names it, and which cited pages are on the project's site. | |
| models | No | The API models to ask. Defaults to the four, chat_gpt, claude, gemini and perplexity, as Prompt Explorer asks them; add grok, deepseek or muse_spark to ask them too (asked directly, so their answers can differ from the app; deepseek and muse_spark answer from what the model learned, without searching the web); [] asks only the apps. | |
| prompt | Yes | The prompt to ask, as a person would type it. | |
| country | No | The country the answers are asked from. Defaults to the United States. Gemini, Grok, DeepSeek and Muse Spark take no country. | US |
| versions | No | The version to ask an API model at, keyed by the model, from Prompt Explorer's picker: chat_gpt gpt-6.1-sol, gpt-6-sol, gpt-6-astra, gpt-6-luna, gpt-5.6-terra, gpt-5.6-sol, gpt-5.6-luna, gpt-5.5, gpt-5.4, gpt-5.4-mini; claude claude-opus-5-5, claude-sonnet-5-5, claude-opus-5, claude-sonnet-5, claude-opus-4-8, claude-sonnet-4-6, claude-haiku-5-5, claude-fable-5-1, claude-fable-5; gemini gemini-4-argon, gemini-3.8-flash, gemini-3.6-flash, gemini-3.1-pro-preview; perplexity sonar-reasoning-pro, sonar-pro, sonar; grok Grok 4.7 (its only version, so leave it out); deepseek DeepSeek V4.1 Flash (its only version, so leave it out); muse_spark Muse Spark 1.3 (its only version, so leave it out). A model left out is asked at the workspace's default version from Settings, else the built-in one. | |
| projectId | Yes | Required. The Visibility Genies project this call works in and is billed to. Call list_projects to find one. | |
| webSearch | No | Whether the API models search the web to answer. Defaults to true, as the page asks. Perplexity and Grok always search, and DeepSeek and Muse Spark never do. | |
| refreshApps | No | Also buy a new answer from each app whose saved one is out of date (after 7 days) or empty, as Prompt Explorer's Refresh does. Defaults to false. An API model's saved answer never goes out of date and is never bought again here; Prompt Tracking buys new ones on a schedule. | |
| approvedMaxCostUsd | No | The most, in USD, the user approved for this call after seeing its estimate and maximum. Only needed after a call was refused for costing more than the workspace's agent spend limit; never set it without asking the user. |
Output Schema
| Name | Required | Description | Default |
|---|---|---|---|
| cost | No | ||
| meta | No | ||
| bought | Yes | ||
| prompt | Yes | ||
| answers | Yes | ||
| unreadable | Yes | ||
| savedResultId | No |