What is a decodable book generator?

A decodable book generator is software that produces new decodable books on demand inside a defined phonics scope. Storytime's generator anchors to a specific lesson in one of the supported curricula (UFLI Foundations, Wilson Fundations, Amplify CKLA, IMSE Orton-Gillingham, LMW, Storytime AI) and produces a brand-new book — narrated, illustrated, decodability-scored — in 15-30 seconds.

How is this different from a generic AI 'story generator'?

Generic story generators ignore phonics constraints. A decodable book generator hard-caps the phonics patterns and heart words to what the student has been explicitly taught at the chosen lesson. The teacher sees the decodability percentage as a number — not a marketing claim — and the system blocks generation outside the eligible scope. The output is instructional material, not a story toy.

Can the AI generate a book that uses untaught phonics?

No. The phonics caps come from the lesson the teacher selects. If a pattern hasn't been taught at that lesson in the chosen curriculum, the model is structurally prevented from producing words that depend on it. Decodability percentage is computed honestly against the student's accumulated phonics inventory.

Do the illustrations support guessing from pictures?

No. Illustrations are story-supporting — they show the setting, characters, and mood — but they do not telegraph the next word's identity. This is a deliberate constraint to prevent the picture-cueing pattern that level-readers train. Students practice decoding, not predicting.

Who can use the generator — teachers only, or students?

Teachers always. Parents on the family plan. Students can generate inside a teacher-defined sandbox — pre-approved curriculum scope, pre-approved illustration styles, pre-approved themes. Students never get unconstrained access; the teacher's curriculum stays the source of truth.

How fast is generation?

15-30 seconds end-to-end — text, narration, and illustrations. The book appears in the library immediately and is available to the student or class you generated it for. Faster than finding a comparable pre-built book in most catalogs.

What's the cost per book?

Teacher Plan includes 500 AI credits per year (~100 generated books). School and district plans get a custom credit pool sized to enrollment. Generation cost is folded into the plan — no per-book surcharge.

AI decodable book generator

Generate a lesson-aligned decodable in 30 seconds.

Pick a curriculum. Pick a lesson. Pick the audience. Storytime generates a brand-new decodable book inside the scope — narrated, illustrated, and ready to read in seconds. Anchored to the exact phonics patterns your student has been taught. Nothing untaught sneaks in.

Start a free classroom

See all three decodable tiers →

15-30 seconds end-to-end · 500 AI credits per year on Teacher Plan

How it works

Four steps. The first three are the teacher's job; the fourth is the AI's.

Pick the curriculum + lesson

Choose which of the supported curricula you're teaching and which specific lesson. The generator's phonics caps come from the lesson — no untaught patterns can appear in the book.

Choose the audience

K-2 standard decodable, or Hi-Lo (grades 3-12 age-respectful). Same phonics constraint, different plot and tone.

Set the controls

Character, illustration style, story length (7 / 14 / 21 paragraphs), optional plot prompt. Optional teacher-supplied story vocabulary the model can spend sparingly.

Generate

15-30 seconds later: a brand-new decodable in the library with narrated audio, illustrations, honest decodability percentage, and a lesson-alignment tag. Available immediately to the student or class you generated it for.

Controls

Seven inputs. Pick what matters, leave the rest at defaults.

The first two — curriculum and lesson — are required. Everything else has a sensible default. Power-user teachers customize all seven; most teachers set two or three.

Curriculum + lesson: UFLI L42, Wilson Step 2.3, IMSE Module 4 Lesson 7, etc. — the lesson's phonics inventory caps the generator.
Audience age: K-2 standard decodable, or Hi-Lo grades 3-12 with age-respectful plots, settings, themes.
Character: Pick from the student's known characters or a class character. Image style stays consistent across that student's books.
Illustration style: Pixar-3D (default), watercolor, anime, hand-drawn — set per student or per book.
Length: Short (7 paragraphs), Standard (14), Extended (21). Hi-Lo defaults to Extended for older readers.
Plot prompt: Optional one-line topic — 'a story about a kid learning to skateboard,' 'a girl visiting her grandmother on a farm.' Model interprets within phonics constraints.
Story vocabulary: Optional 1-3 teacher-supplied story-specific words the model is allowed to introduce sparingly. Lets you push beyond the lesson's vocab when there's instructional reason.

Quality bars we don't relax

Generation that respects the phonics constraint — not just the plot.

Every generated book has to clear five bars. We don't relax any of them to flatter the catalog or accommodate a harder-to-write plot.

Phonics pattern caps

The lesson's accumulated phonics inventory is hard-capped. The model cannot reach for an untaught pattern just because the plot would be better.
Decodability percentage

Every generated book is scored against the student's phonics inventory. The number shown is honest — not relaxed because the book is older-themed or harder to write.
Heart-word constraint

Only heart words the student has been taught can appear. Untaught high-frequency irregulars are blocked at generation.
Narration generated alongside

Every book ships with fluent modeled narration — same voice as the student's other books unless overridden.
Illustrations support but don't enable guessing

Images are story-supporting, not picture-cued. They don't telegraph the next word so kids practice decoding, not predicting.

See what's possible

Same generator. Different students. Different art styles.

Fourteen students. Fourteen art styles — the exact set teachers pick from in the app. Every classroom has different kids, and every kid deserves to see themselves in the book they're reading. The generator hands teachers the controls; the AI does the rendering.

Sofia · age 5

Pop Art

Maya · age 9

CGI

Diego · age 7

Watercolor Dream

Ethan · age 6

Vintage Storybook

Mateo · age 4

Retro Cartoon

Anika · age 4

Dreamy Pastel

Noah · age 4

Bold Graphic

Kai · age 8

Japanese Woodblock

Fatima · age 10

Photo Realistic

Liam · age 5

Anime

Yuki · age 11

Crayon Drawing

Priya · age 8

Animated Movie

Aaliyah · age 8

Comic Book

Tariq · age 6

Coloring Book

The same character can be rendered in any of these styles. Teachers set the default per student; students never have to look at art that doesn't feel like them.

When teachers use the generator

Four scenarios that recur every week — solved in a minute, not a planning period.

Small group

Five students all stuck on /sh/. The library only has two /sh/-heavy books.

Generate three more on the spot — different plots, different characters, same /sh/ density. The group rotates through five books over five days without you having to scavenge from external sites.

Time-to-book: ~90 seconds

Reluctant reader / engagement

A student who won't open anything — until the hero looks like them.

Generate the lesson's decodable starring a character your student picks: their name, their interests, an illustration style they chose. Representative "mirror" books are why classrooms that add them see student reading time rise by about 4 hours per week (First Book, Diverse Books Impact Study). Same phonics scope — a far more willing reader.

Time-to-book: ~30 seconds

Hi-Lo / striving older reader

A 4th-grade student on short vowels who refuses the K-1 library.

Generate a Hi-Lo decodable with the same phonics constraint and an age-appropriate plot. The student practices decoding without the dignity cost.

Time-to-book: ~30 seconds

Intervention loop

Student stuck on the prefix lesson. Needs extra practice with no progression.

Generate a decodable focused on the specific prefix the student is on. Assign as a 'Special Mission' practice activity that becomes their next-session top priority.

Time-to-book: ~30 seconds

Homework

Homework needs a book the parent can guide the child through — at the exact phonics level.

Generate a parent-friendly decodable for the lesson the student is on. Parent gets the same book in the family-side library with parent-facing guidance.

Time-to-book: ~30 seconds

What it costs

500 AI credits per year on Teacher Plan — about 100 generated books.

Credit cost is folded into the Teacher Plan price ($29.99 first year, $59.99 renewal). No per-book surcharge. District plans get a custom credit pool sized to enrollment.

See the pricing detail

The instructional integrity contract.

We treat generation as instructional material, not as creative writing software. The teacher's curriculum stays the source of truth. The phonics scope cannot be overridden by the model. The decodability percentage cannot be flattered. Heart-word inventory is enforced at the lesson level. Generation is lesson-gated — students cannot use the generator outside the scope of their current lesson. The teacher is always in control.

FAQ

What the decodable book generator does (and doesn't).

What is a decodable book generator?: A decodable book generator is software that produces new decodable books on demand inside a defined phonics scope. Storytime's generator anchors to a specific lesson in one of the supported curricula (UFLI Foundations, Wilson Fundations, Amplify CKLA, IMSE Orton-Gillingham, LMW, Storytime AI) and produces a brand-new book — narrated, illustrated, decodability-scored — in 15-30 seconds.
How is this different from a generic AI 'story generator'?: Generic story generators ignore phonics constraints. A decodable book generator hard-caps the phonics patterns and heart words to what the student has been explicitly taught at the chosen lesson. The teacher sees the decodability percentage as a number — not a marketing claim — and the system blocks generation outside the eligible scope. The output is instructional material, not a story toy.
Can the AI generate a book that uses untaught phonics?: No. The phonics caps come from the lesson the teacher selects. If a pattern hasn't been taught at that lesson in the chosen curriculum, the model is structurally prevented from producing words that depend on it. Decodability percentage is computed honestly against the student's accumulated phonics inventory.
Do the illustrations support guessing from pictures?: No. Illustrations are story-supporting — they show the setting, characters, and mood — but they do not telegraph the next word's identity. This is a deliberate constraint to prevent the picture-cueing pattern that level-readers train. Students practice decoding, not predicting.
Who can use the generator — teachers only, or students?: Teachers always. Parents on the family plan. Students can generate inside a teacher-defined sandbox — pre-approved curriculum scope, pre-approved illustration styles, pre-approved themes. Students never get unconstrained access; the teacher's curriculum stays the source of truth.
How fast is generation?: 15-30 seconds end-to-end — text, narration, and illustrations. The book appears in the library immediately and is available to the student or class you generated it for. Faster than finding a comparable pre-built book in most catalogs.
What's the cost per book?: Teacher Plan includes 500 AI credits per year (~100 generated books). School and district plans get a custom credit pool sized to enrollment. Generation cost is folded into the plan — no per-book surcharge.

About the authors

Written and edited by the Storytime AI founding team.

Brian Carlson

Co-founder & CEO

Co-founder and CEO of Storytime AI. Leads the company from Baltimore, building a literacy platform that meets every reader where they are — anchored to the Science of Reading.

Scott Quinlan

Co-founder & CTO

Co-founder and CTO of Storytime AI. Owns engineering, product infrastructure, and the agentic growth pipeline — from the platform's AI generation engine to the structured-literacy content surface district leaders evaluate.

Kate Dwyer

Co-founder & CMO

Co-founder and Chief Marketing Officer at Storytime AI. Translates Science-of-Reading research and product capability into language teachers, parents, and district leaders can act on. Based in the Washington DC–Baltimore area.

Generate the book your student needs for tomorrow.

Pick the curriculum. Pick the lesson. Pick the audience. Have it in the library before the bell rings.

Start a free classroom

See the full customization layer →