API Reference

Generate realistic made-up personas with detailed psychological profiles. IMPORTANT: If the user specifies a number in their request (e.g., 'six engineers', 'three doctors'), use that number for the count parameter. We will suggest clarifications for broad demographics or generate immediately for specific ones.

Browse all available made-up personas with optional demographic filtering. Supports pagination using limit (max results per page) and offset (number to skip) parameters

Show complete background information for a specific made-up persona

Search made-up personas using natural language. Express any search criteria naturally - demographics, personality traits, professions, age ranges, locations, etc. (e.g. 'creative engineers in startups', '25-year-old designers from Europe', 'introverted marketing managers')

Create a new saved made-up persona group with specified made-up personas

List all saved made-up persona groups for the authenticated user

No parameters required.

Get detailed information about a specific saved made-up group including all made-up personas

Update a saved group: add/remove made-up personas, rename, or update description

Delete a saved made-up persona group (this will not delete the made-up personas themselves)

Have a conversation with a specific made-up persona who responds in character according to their psychological profile and background. IMPORTANT: Pass the user's exact question/dialogue unmodified - do not rephrase, summarize, or interpret their words. OPTIONAL: Set includeRating=true to generate numeric rating distributions (1-5 scale with probabilities). Optionally specify scaleType (purchase_intent, importance, satisfaction, likelihood, agreement, sentiment) to match your question construct.

Ask a question to a group of personas. Use groupId for exact matches, groupName for smart matching to groups, or personaIds for direct specification. IMPORTANT: Pass the user's exact question/dialogue unmodified - do not rephrase, summarize, or interpret their words. OPTIONAL: Set includeRating=true to generate numeric rating distributions for each persona (1-5 scale with probabilities). Optionally specify scaleType (purchase_intent, importance, satisfaction, likelihood, agreement, sentiment) to match your question construct.

Continue an existing threaded conversation with a made-up persona

Continue an existing group conversation with the same made-up personas or a subset of them

Get detailed psychological analysis explaining why a made-up persona responded the way they did to your question. Use the messageId from any conversation response

Conduct structured interviews with personas. Returns text responses with automated sentiment analysis converting each response into 1-5 rating distributions and aggregate statistics (mean, distribution) per question. Match scaleType to question construct (satisfaction, likelihood, etc.). Returns Task ID - poll with task_status, retrieve with task_result.

Perform statistical and thematic analysis on completed interviews. Provides descriptive statistics, group comparisons (t-test/ANOVA), outlier detection, and optional qualitative theme extraction for demographic segments. Requires Pro+ subscription. Use approve=false for preview/cost estimate. Returns Task ID - poll with task_status, retrieve with task_result.

Retrieve detailed data for analysis drill-down. Fetch full interview transcripts by interview_ids, specific exchanges by exchange_ids, all exchanges for conversation_ids, or persona profiles by persona_ids. Use exchange_ids from analysis outliers to investigate, or filter by codes to explore thematic clusters. Requires Pro+ subscription.

Check the status and progress of a long-running task. Returns current status (queued/running/completed/failed/cancelled), progress percentage (0-100), and optional progress message. Use this to poll tasks that return a taskId

Retrieve the final result of a completed long-running task. Returns the full result data if status is 'completed', or an error if the task failed. Use task_status first to check if ready

Cancel a running long-running task (cooperative cancellation). The task will stop at the next cancellation check point. Only works for tasks with status 'queued' or 'running'

List recent long-running tasks for the current user. Shows task summaries ordered by creation time (newest first). Filter by status (queued/running/completed/failed/cancelled) and limit results (max 50). Use this to see your task history and find taskIds

View available subscription plans and pricing. No parameters required

No parameters required.

View current subscription plan and usage statistics. No parameters required

No parameters required.

General subscription management. Params: action={upgrade|portal}, tier={plus|pro|max}, currency={usd|eur|gbp}. IMPORTANT: Before calling, infer currency from conversation (user location/preferences) or explicitly ask the user which currency they prefer

Upgrade to a higher subscription tier. Params: tier={plus|pro|max}, currency={usd|eur|gbp}. Returns Stripe checkout URL for payment. IMPORTANT: Before calling, infer currency from conversation (user location/preferences) or explicitly ask the user which currency they prefer

Access Stripe billing portal to manage subscription (paid users only). Update payment method, view invoices, cancel subscription. No parameters required

No parameters required.

Learn about I Made It Up: how made-up personas work, getting started guide, workflows, and important terms

No parameters required.

Submit feedback in your own words about your experience. Describe what happened, what you expected, or suggestions for improvement