[ Use cases ] # Jev for data teams: what it can do and what it costs How data teams can use Jev, TypeSafe AI's decision model: 3 jobs, the question for each and the list-price cost per 1,000. Checked 1 November 2026. Published 1 November 2026By Jev Hub [Back to use cases](https://jevhub.io/use-cases) [ TLDR ] - Records that should match do not, documents need sorting into long category lists, and values hide in messy text. - Jev, TypeSafe AI's decision model, can take on these jobs for data teams: Decide which records to merge, which to keep apart and which a person should check (1.2¢ per 1,000 pairs); Pick one of dozens of categories, and fall back to a broader one when unsure (11¢ per 1,000 documents); Find the right value in messy text without letting a model make one up (2.9¢ per 1,000 documents). - Each is one request per item that returns a typed value with a probability, so your own code decides what happens next. - It does not write replies, so it works alongside a person or a language model, not instead of one. ## The jobs, the questions and the cost for data teams | | What Jev is asked | Cost per 1,000 (list price) | How good a fit | | --- | --- | --- | --- | | Decide which records to merge, which to keep apart and which a person should check | Score, Noul, 4 questions per pair | About 1.2¢ per 1,000 pairs | Good fit | | Pick one of dozens of categories, and fall back to a broader one when unsure | Choice, 1 question per document | About 11¢ per 1,000 documents | Good fit | | Find the right value in messy text without letting a model make one up | Choice, 2 questions per document | About 2.9¢ per 1,000 documents | Good fit | Costs are the calculator's estimate at each use case's typical volume, at Jev's list price of $0.042 per million input tokens (read from TypeSafe's models page on 1 November 2026). They are estimates, not measured bills. ## What can data teams use Jev for? Jev, TypeSafe AI's decision model, answers a question about text with a value your software can use: one option from a list, a score, or a yes/no probability. That suits the small, repeated decisions in the middle of this work. These are the jobs on this site that fit data teams best. Each has its own page with the questions to ask, the pitfalls and a worked example. - **Decide which records to merge, which to keep apart and which a person should check.** For each candidate pair, one Score question with three levels (different, possibly related, same) decides whether to leave the records apart, ask a person or merge them, and a few yes/no questions in the same request show which fields disagree. - **Pick one of dozens of categories, and fall back to a broader one when unsure.** In TypeSafe's cookbook, 60 company annual reports were classified into 75 industry groups with one Choice question each. - **Find the right value in messy text without letting a model make one up.** For values such as emails, phone numbers and amounts, a pattern match finds the candidates and Jev picks the one the question asks for, so it cannot invent a value. - [Jev for matching duplicate records](https://jevhub.io/use-cases/record-matching) - [Jev for classifying documents into many categories](https://jevhub.io/use-cases/document-classification) - [Jev for extracting dates, emails and amounts](https://jevhub.io/use-cases/data-extraction) ## What does it cost? Jev lists $0.042 per million input tokens with output free, so the cost depends on how much text you send and how many questions you ask about it. At the typical volumes the use-case pages assume, the calculator estimates: - **Pairs (matching duplicate records):** about 1.2¢ per 1,000, at 4 questions and about 120 tokens each. - **Documents (classifying documents into many categories):** about 11¢ per 1,000, at 1 question and about 1500 tokens each. - **Documents (extracting dates, emails and amounts):** about 2.9¢ per 1,000, at 3 questions and about 500 tokens each. - [Cost calculator](https://jevhub.io/calculator) - [Jev pricing and real runs](https://jevhub.io/economics) ## Where to start A plain way to begin, in order: - Pick one cleanup that is blocking a report, such as deciding whether two records are the same thing. - Turn it into a Score or Choice question with the levels or categories your team already uses. - Use the confidence value to split the output into answers you trust, answers a person should check, and a broader fallback. - Write the checked cases back as labelled data, so the next run starts better. The order to try Jev in. Our suggested order, which follows TypeSafe's advice to test on labelled data of your own before relying on the answers. - [How to use Jev](https://jevhub.io/how-to-use-jev) ## When Jev is the wrong tool for data teams For extraction, let Jev choose among candidates your code found rather than ask it to produce a value. That is how TypeSafe's cookbooks keep it from inventing one. From the use-case pages, Jev is not the right choice when: - Records share a reliable ID. Match on the ID in code. - Every document must get a precise label with no review. Plan for the unsure half. - The document has a fixed layout. A parser is cheaper and exact. - [What Jev can't do](https://jevhub.io/jev-limitations) ## Our read Data cleaning is the place where the confidence value earns its keep. Being able to say 'half of these are safe to merge, half need a person' is worth more than a slightly higher accuracy number. The figures above are list-price estimates, not measured bills, and the accuracy claims on the linked pages are the published runs' own. Test Jev on a labelled sample of your own data before you rely on it. - [What is a decision model?](https://jevhub.io/what-is-a-decision-model) ## Common questions Yes. For each candidate pair, one Score question with three levels (different, possibly related, same) decides whether to leave the records apart, ask a person or merge them, and a few yes/no questions in the same request show which fields disagree. In TypeSafe's cookbook on 450 beer-catalogue pairs, 360 were left unlinked, 40 merged and 50 sent to a curator. Yes, and its confidence tells you which answers to trust. In TypeSafe's cookbook, 60 company annual reports were classified into 75 industry groups with one Choice question each. A confidence cutoff of 0.9 split them in half: the confident half was right 90% of the time, the other half 40%, and reporting that half one level up the hierarchy raised it to 70%. Yes, in a specific way: Jev chooses, and code does the rest. For values such as emails, phone numbers and amounts, a pattern match finds the candidates and Jev picks the one the question asks for, so it cannot invent a value. For dates, Jev reads the parts the text names (month, day, year or 'next Thursday') and code builds and checks the date. Both are TypeSafe cookbooks. Jev lists $0.042 per million input tokens with output free. At typical volumes the calculator estimates 1.2¢ per 1,000 pairs, 11¢ per 1,000 documents, 2.9¢ per 1,000 documents. These are list-price estimates, not measured bills. No. According to TypeSafe, System One models such as Jev return typed answers and probabilities and do not write replies, produce code or explain their reasoning. Use a person or a language model for the writing. Pick one cleanup that is blocking a report, such as deciding whether two records are the same thing. Use the confidence value to split the output into answers you trust, answers a person should check, and a broader fallback. ## Sources Written by Jev Hub from the sources below and checked against them on 1 November 2026. Where a figure is a company's own claim, the article says so. - [TypeSafe docs: System One (read 1 November 2026)](https://docs.typesafe.ai/concepts/system-one) - [TypeSafe docs: Primitives, the question types (read 1 November 2026)](https://docs.typesafe.ai/primitives) - [TypeSafe docs: models and prices (read 1 November 2026)](https://docs.typesafe.ai/models) - [TypeSafe cookbook: Knowledge graph entity alignment](https://docs.typesafe.ai/cookbooks/entity_alignment) - [TypeSafe cookbook: Classification using confidence](https://docs.typesafe.ai/cookbooks/classification_using_confidence) - [TypeSafe cookbook: Pre-parsed value extraction and date extraction](https://docs.typesafe.ai/cookbooks/pre_parsed_value_extraction_cookbook) ## Related content - [What is Jev?](https://jevhub.io/what-is-jev) - [How to use Jev](https://jevhub.io/how-to-use-jev) - [Jev pricing](https://jevhub.io/economics)