iAsk

Free lists unlimited daily searches on the standard, non-Pro model, with limited ads. It also includes the writing tool, the AI Summarizer for URLs and pasted text, search history and a general AI assistant. The Pro model, file uploads to the summarizer, the Document Analyzer and the video tutor are not part of it.

The free plan card also lists Companion AI, described by the provider as 24/7 mental health support. A chatbot is not a clinician, so anyone in distress should contact a local crisis line or a qualified professional rather than rely on it.

Pro caps Pro-model searches at 300 a day, which is generous for one person studying but finite for someone running many follow-ups on a long project. The pricing page does not say whether searches beyond the cap fall back to the standard model or stop until the next day.

The terms also let the operator enforce quotas and throttle usage at its sole discretion, with or without notice. They forbid sharing one account among several people to avoid fees, so a household or study group needs separate accounts rather than one shared Pro login.

The provider promotes Pro with benchmark results, but the figures change from page to page: 85.85% on MMLU-Pro and 78.28% on GPQA on the pricing page, 86.55% on MMLU on the homepage, and 93.89% on MMLU with 90.1% on TruthfulQA in the student FAQ. These are self-reported scores on different tests and do not show how an answer to your question will turn out.

The terms are more cautious than the marketing. They make no warranty on the accuracy of output and say it is for informational purposes only. Treat each answer as a pointer to its sources, and quote the original document rather than the generated summary.

The homepage says iAsk answers questions without storing student data and claims no personal data or conversation harvesting. The privacy policy, which governs the service, is broader: it says chat inputs and transcripts may be collected, monitored, stored and recorded by iAsk and its third-party vendors, along with device and usage data.

The same policy allows targeted advertising through third-party ad companies, sharing with advertisers and investors for business analysis, and sending search queries to the YouTube Data API. Keep names, grades, health details and unpublished work out of your questions, since the policy, not the homepage slogan, sets what can happen to them.

The student page offers one year of Pro with a .edu or .ac email and says no credit card is required. It is described as a limited-time offer without an end date, and the page does not explain what happens when the year ends. Without stored card details, an automatic charge is unlikely, but the terms do not say how the account is treated afterwards.

The student FAQ quotes the regular price as typically $9.99 a month, while the pricing page shows $9.95. Check the price on the checkout screen if you decide to continue after the free year.

The API is a separate product billed per response rather than per month: $1.20 per 1,000 responses on Basic, $6 on Regular and $12 on Advanced, the level that includes the iAsk Pro model. Regular adds a question mode, and all tiers are pitched at developers building search or content tools.

• Use Pro for your own reading and study within 300 Pro searches a day.
• Use the API when another application needs answers programmatically.
• Budget the API from expected responses, not from the Pro subscription price.

1. Record the date, plan and response detail level used.
2. Ask each question once and save the answer with its citations.
3. Open every cited link and find the sentence that supports the answer.
4. Mark each answer as supported, partly supported or unsupported by its own sources.
5. Repeat one question to see whether the answer or sources change.

Accept the result if every answer matches the known fact and at least one citation contains the supporting statement. Reject answers whose links do not contain the claim or point to pages that fail to load. Leave the observed scores blank until the questions are run.

For cited research with source filters and follow-up threads, compare Perplexity. For answers drawn only from academic papers, see Consensus or Elicit.