StoryHelm is a continuity checker built for one reader: the indie author several books into a series. You bring the manuscripts together, it reads each book for contradictions, and a cross-book check you start compares what the books record about the same characters and events, leaving each potential conflict for you to judge. Its homepage states the position as "It reads. It never writes your prose.", with two disclosed exceptions covered below.
This review assesses the vendor's current public pages, pricing and privacy notes rather than hands-on testing. That is not only a matter of method: StoryHelm has not opened to the public yet. Its pricing page says public launch follows a founding cohort, whose members get access in waves, and the free self-audit signup is the founding list.
The quick verdict
The most tightly aimed product in this group at the problem series authors actually describe, with evidence shown before any verdict and a model provider named plainly. The limits are practical: you cannot simply buy it today, and the plan built for series sits at the expensive end of the category.
Prices*
| Plan | Billed annually | Billed monthly | Built for |
|---|---|---|---|
| Writer | USD 14 per month | USD 19 per month | One book at a time, three projects |
| Author | USD 35 per month | USD 44 per month | Single-book craft, ten projects |
| Series | USD 59 per month | USD 74 per month | Unlimited books, a canon for each one gathered in a Series Atlas |
| Studio | USD 199 per month | Annual only | Five co-author seats, real-time co-editing, 5,000 analysis credits a month |
Two things sit outside the plans. The Manuscript Audit is USD 99 once: StoryHelm reads a single book and returns a findings report you keep, arranged directly with the founder while the cohort runs. Translation Studio is a separate per-book service, from USD 199 per language for the first tier of European languages and from USD 349 for the second tier. A free twelve-check Series Continuity Self-Audit arrives as a PDF by email, and a 14 day free trial is promised at public launch.
What it does well
- Built around the series problem. The flagship plan builds a canon for every book, which you can confirm or correct, and gathers them in a Series Atlas. A cross-book check compares a character's recorded appearance and distinguishing marks between books, their voice where each book holds 20 or more lines of their dialogue, and, between two fully read books, the year, month, season or length given for the same event.
- Evidence first. A cross-book finding names both books and shows what each one records, and whether a change is a revelation or a slip is left to you, which is the right way round for deliberate reversals.
- The checker only reads. The vendor is explicit that it will not draft or rewrite your prose; it flags the discrepancy and you make every change.
- Plain data terms. Analysis runs on Claude, by Anthropic; the vendor states that neither company trains a model on your manuscript, and says openly that Anthropic keeps API inputs and outputs for a limited window before deletion.
- A useful free resource. The self-audit is a working checklist of twelve error types, usable on your own shelf without uploading anything.
Where StoryHelm stops, and where The Novel Kit does
- Availability. Until public launch, access runs through the founding cohort.
- A narrow cross-book check. Between books it compares appearance, voice and the timing of the same event, and nothing else. It does not compare names, so a renamed character is filed as a new one; it does not compare world rules, relationships and ranks, places and distances, or who knew what; and there is no master series timeline. World-rule breaks are flagged inside a single book only. The check runs when you start it and confirm its credit cost, not automatically.
- No quoted passages between books. A cross-book finding shows what each book records rather than the lines themselves, so you go back to both books to find them.
- Storage. Files stay under your account until you delete them, and deletion or export is by email, answered within about 30 days.
- Not every part only reads. The never-writes promise covers your prose, and the vendor discloses two exceptions. Ship Dock drafts pitch material about the book, such as a query letter, synopsis, logline or back-cover blurb, for you to edit. Translation Studio produces machine-translated editions of a finished book, a whole new text. This index therefore records the product as offering optional generative AI.
- Cost at series scale. The Series plan on annual billing comes to USD 708 a year.
- Judgement stays with you. A flag is a model's reading of the books, and it has to be verified against the text.
| Question | StoryHelm | The Novel Kit |
|---|---|---|
| How it checks | A language model reads each book, then compares what the books record | Fixed rules and pattern matching, no AI at any point |
| What it can catch | Contradictions of meaning. Across books: appearance, voice, and the date or length of the same event. Within one book, broken world rules and drifting dates too | Literal drift: names, invented terms, spelling and style decisions |
| What happens to the manuscripts | Kept in your account until you delete them | Transiently processed for the scan; never stored |
| How you pay | Monthly or annual plans, or a one-off audit | A pass or an annual plan, with unlimited scans |
| Can you start today | Through the founding cohort | Yes |
The two tools answer different questions. StoryHelm asks whether book three describes a character or dates an event differently from book one. The series check in The Novel Kit asks whether the books agree with themselves on the page: a name close enough to another to be a slip, a term written another way, a term the earlier book established and this one drops. It will not tell you a character's eyes changed colour, and says so. The overlap is small in the other direction too: StoryHelm does not compare names between books, which is where The Novel Kit starts. Run the literal pass first and a model's flags are easier to trust, because spelling noise is already gone.
A practical next step
Download the free self-audit and run it by hand on two of your books this week; it costs nothing and tells you which of the twelve error types your series is prone to. If you then join the cohort, test the checker on a slip you already know is there, following the guide to verifying an AI continuity report.
Sources checked
- StoryHelm
- StoryHelm pricing
- StoryHelm on what its continuity check does
- StoryHelm privacy policy
- StoryHelm on translation pricing
- Best tools for checking a series for continuity
What a writing app usually does not check
Drafting apps, codexes and wikis hold what you put into them. These checks need the finished manuscript read as a whole, and this is what The Novel Kit does about each.
| What slips through | What The Novel Kit does |
|---|---|
| One invented name spelt two ways: Velora in chapter two, Vellora in chapter nine | Compares every name with every other and pairs the near-identical ones, with a count of each spelling. Duplicate and alias checker |
| Two different characters whose names look alike at a glance | Flags crowded initials, shared endings and names of the same length. Name shapes |
| Your own terms capitalised or hyphenated more than one way | Counts every form each term takes across the book. Capitalisation and hyphenation checkers |
| Terms used again and again and never explained | Lists invented terms with three mentions or more and no defining sentence anywhere in the book. Unexplained terms |
| A sequel that drifts from the earlier books | Checks book two against book one for names spelt two ways, dropped terms and house style that moved. It compares the words, not the story facts. Series check |
Rule-based, with no AI reading the manuscript, which is transiently processed and never stored. The free preview needs no account.
* Prices and plan details are public list prices taken from each product's own pages, exclude tax unless stated, and were last checked in 2026. Vendors change them without notice, so confirm the current figure in the seller's checkout before buying.
Frequently asked questions
What is StoryHelm?
A continuity checker for indie authors writing a series. It reads each book for contradictions, and a cross-book check you start compares a character's appearance and voice between books and the date or length of the same event. It shows what each book records and leaves the decision to you. It does not compare names or world rules between books.
Can I buy StoryHelm now?
Not on the open market. The pricing page says public launch follows a founding cohort, whose members get access in waves, and the free self-audit signup is the founding list.
What does StoryHelm cost?
Writer is USD 19 per month, or USD 14 per month billed annually. The Series plan is USD 74 per month, or USD 59 billed annually, and a one-off Manuscript Audit is USD 99.
Does StoryHelm use AI?
Yes. Analysis runs on Claude, by Anthropic, and the checker does not write your prose. Two parts do generate text: Ship Dock drafts pitch material such as a query letter or synopsis for you to edit, and Translation Studio produces machine-translated editions.
How is it different from The Novel Kit?
StoryHelm uses a language model to look for contradictions of meaning, does not compare names between books, and keeps your files. The Novel Kit stores no manuscript and uses fixed rules to report names, terms and style decisions that differ between books.