Help centre · 98 answers
Questions, answered.
Everything people ask about Induce AI in one place: scripts and continuity, shot control and models, dubbing, product photos, templates, pricing and who it is for. 98 answers across 17 topics, each linked to the page it comes from.
Getting started
About Induce AI
What Induce AI is, who it is for and how it works.
Induce AI is an AI filmmaking workspace for studios and production organizations. It connects a brief or screenplay with scene planning, visual references, generated clips, and a reviewable final cut.
Test a representative sequence with recurring characters, wardrobe, props, and locations. Review the shots together, record drift and retakes, and use the result to plan production. Visual references support the workflow; creative review is part of the process.
Induce AI works with external generation models within its filmmaking workflow. Ask our team for current model and quality-setting availability for your intended production; versions and exposed capabilities can change.
Generation and retries are subject to your plan’s usage terms. Enterprise agreements define included usage, retry allowances, and overage rates. Include retakes, review, editing, and finishing when comparing production costs.
Induce AI’s Terms say customers retain ownership of outputs from their inputs and have commercial usage rights, subject to the Terms and applicable third-party rights. Review the Terms for your production requirements.
Induce AI breaks a screenplay or brief into a scene plan split into clips, attaches character, location, and prop references, generates each clip with a supported image or video model, and assembles a reviewable cut in one project. See the screenplay-to-film workflow.
Verification models compare the story state planned for a clip with sampled frames from the generated take. Each check returns pass, fail, or unknown; missing evidence stays unknown rather than counting as a pass. Reviewers add notes before a take is kept. How story-aware verification works.
Shared references and the planned story state carry from clip to clip, and takes are reviewed in sequence. In Signal at Dawn, human review flagged a harbor warning lamp that was lit one clip earlier than the plan. Explore AI video continuity.
The Studio workspace is built around multi-scene narrative: scenes, recurring character and location references, and continuity review across the cut. Scope, volume, and delivery for a series are agreed in a pilot.
Through a sales-led production pilot scoped to a representative script, your acceptance criteria, volume, and delivery requirements. Plan an enterprise pilot.
Induce AI is an AI filmmaking platform for studios, media platforms, and agency networks. It turns a screenplay into a scene plan, references, generated clips, and a reviewable cut, and keeps the story consistent from one shot to the next.
Rhapsody is the reasoning and verification engine inside Induce AI. It records what each shot must show and checks generated takes against it. Each check is a match, a mismatch, or unclear; unclear is never counted as a match.
A video generator makes one clip from one prompt. Induce AI works across the whole scene: it breaks down the script, plans the shots, generates with supported image and video models, and checks takes against the story before the next shot depends on them.
Directors and producers developing narrative work, studios and media teams producing at volume, and agency networks delivering client work. Teams usually start with a pilot scoped to a representative scene.
Pricing and credits
Cost per finished minute, monthly credits and Enterprise plans.
Induce AI typically lands between $6 and $30 per finished minute at 720p–1080p, about $16 on average, depending on the model, resolution, shot count, and retakes. 4K output costs more. Traditional production commonly runs $5,000 to $40,000 per finished minute. Measure your own cost on a pilot scene before you plan a full production.
Each plan includes a monthly credit allowance that is spent on generation across the supported video and image models. The amount a shot uses depends on the model, duration, and quality setting you choose for it.
Talk to us when you need production volume beyond the self-serve plans, SSO/SAML, VPC deployment, API or webhook access, or custom export and delivery requirements. Availability and timing for these are confirmed before rollout.
Product
Script to video
Turning a script or brief into scenes, shots and a cut.
Script-to-video AI connects a source script with scene and shot planning, visual references, generated takes, and an assembled cut for review. In Induce AI, those stages are organized within a project.
The current script upload workflow accepts a PDF screenplay. Use a text-based PDF, then confirm scene structure and supported import length with a representative sequence before planning a larger production.
The project provides scene and clip-plan views with direction, dialogue, generation status, and saved takes. Review these against the source script and your creative requirements.
Start with a representative sequence. Establish references and acceptance criteria, then measure approved footage, revision effort, and total production costs before expanding the number of scenes or projects.
Studio includes retake and editing workflows. Test a specific creative change during evaluation, compare the new result with the intended references, and review it in the assembled sequence.
Track cost per approved finished minute, usable takes, review time, revision effort, and delivery acceptance. Include generation attempts, retakes, editing, and finishing in the comparison with your current workflow.
AI video continuity
Keeping characters, locations and props the same in every shot.
Continuity means character identity, wardrobe, locations, props, action, lighting, and spatial relationships make sense across the cut. It is assessed across a sequence as well as within each shot.
Named visual references and scene and shot planning establish the intended look and action. Supported workflows can also use context from an accepted earlier take. Review the generated footage together to judge whether the sequence holds.
User-supplied visual material can be saved as named project references for characters, locations, and props. Confirm the selected workflow’s input and likeness requirements before production.
A new scene can require a new visual setup. References should preserve the intended identities and design, while the scene and shot plan establishes the new place, lighting, action, and story state.
Generated results can vary. References establish the intended identity and look; the production team still needs to review details across shots and decide which results need another take.
Use recurring characters and props across different framings in a representative sequence. Agree a review rubric for identity, wardrobe, action, location, and lighting, then record acceptance, retakes, and review effort across the cut.
Shot control
Revising a single shot and describing a retake.
The Studio final-cut view includes clip actions such as Retake and Replace. Select the clip you want to revisit, make the change, and review the resulting cut before delivery.
Give a specific creative note, such as a closer framing, a clearer action, or a different lighting direction. Keep the intended character and setting references available, then inspect how the new result fits the scene.
A new generation can change visual details. Review identity, props, screen direction, timing, and transitions against the surrounding shots. Use the final cut to judge whether the revision works in context.
Discuss a pilot with a real scene and a few concrete revision notes. Compare the original and revised footage in the cut, record the accepted takes and production effort, and agree any additional editing or delivery requirements with our team.
AI models
How generation models fit into the filmmaking workflow.
Generation models create images or video within a broader project. In Induce AI, teams develop the script, shot plan, and visual references, then generate clips and review the assembled cut. Model choice is one part of that production workflow.
Generation options include model families such as Kling, Seedance, Veo, and Nano Banana across image and video workflows. Available versions, quality settings, and access vary by workflow. Confirm the current options for your production during a pilot.
No model choice guarantees continuity. Use character, location, and prop references, inspect generated shots, and review them together in the cut. Evaluate the settings against your own script and acceptance criteria.
Test representative shots with the same brief and references. Compare usable takes, motion, dialogue, continuity, review effort, and the full production cost. Confirm supported inputs, output formats, and usage terms for the selected workflow.
AI product shoot
Flash Shoot: product photos and films from your own product images.
Product Shoot is Induce AI’s AI product photography workspace. Add product references, choose a model when needed, select outputs and an aesthetic, review the estimate, then generate and inspect the results.
Use 1–8 clear reference photos of the same product. Include useful angles and visible details such as construction, material, color, labels, and logos. Upload material your team has permission to use.
Choose packshots, lifestyle product photography, on-model images, and selected product videos. Available shots depend on your setup and account; confirm the shot list and delivery requirements when planning your pilot.
Plan product photos for catalogs and product detail pages (PDPs), then select lifestyle or on-model shots for campaigns. Check your channel’s image and format requirements and approve product details before publishing.
Choose a suggested look or describe your own direction. A look can define the setting, light, palette, wardrobe, and camera treatment. Review the direction, selected shots, and estimated cost before generation.
Checks can report pass, fail, or uncertain findings. Still-image checks cover the image; video checks sample frames. Compare results with your product references and approve them before publishing. Results without a completed check remain unverified.
Product Shoot shows an estimated credit total for the selected shoot before generation. Review the setup, shot list, and current estimate together. Measure review time and any further attempts when evaluating production cost.
Bring a representative product, its references, required image formats, and your approval criteria. Start with a scoped pilot, review the resulting imagery, and discuss production volume and rollout requirements with our team.
MCP and agents
Using Induce AI from AI apps through the MCP server.
A connection that lets Claude, ChatGPT, Cursor and other AI apps make images and video with your Induce AI account, using the Model Context Protocol. Your AI plans the work; Induce AI makes it.
No. It spends the same credits as Induce AI Studio at the same prices. Every job is priced before it runs, and each app has a monthly cap you set when you connect it.
Claude, Claude Code, ChatGPT, Cursor, Codex, Gemini CLI and any app that supports remote MCP servers with OAuth. Apps without OAuth, such as Grok, connect with a personal token from Induce AI Studio.
No. Each app can only spend up to its monthly cap, every job is quoted first, and any single job above your approval limit waits until you approve it in Induce AI Studio.
Into your Induce AI library, in a project named after the app, so you can open, edit and download everything in Studio.
Open Connect in Induce AI Studio and choose Disconnect. The app loses access on its next request.
Rhapsody
The engine that plans each shot and checks every take against the story.
The Project Memory Graph keeps track of who exists in the story, what they wear and carry, and what has happened. Each new shot is planned from that state, and the Coherence Verifier compares the take with the character references before it is approved.
Rhapsody sits above the generators. Shots are described in one shared format, the Shot Spec IR, and handed to whichever supported image or video model in Induce AI suits the shot. Switching models does not change what the shot has to show.
Re-rolling starts each attempt from nothing. When a take misses, Rhapsody records what failed and why, re-plans only that part, and keeps everything that already worked.
Each check is a match, a mismatch, or unclear, with the evidence that supports it. Unclear is never counted as a match, so a person reviews it before the shot moves on.
No. Rhapsody handles the bookkeeping of continuity so the team can spend its time on creative decisions. Directors set the intent and approve the cut; Rhapsody shows the evidence behind each decision.
Rhapsody runs inside Induce AI Studio. You can sign up and open a project, or talk to the team about a pilot scoped to a representative scene from your script.
Who it is for
Filmmakers
Working from a screenplay and revising shots.
Directors, producers, and development teams within studios and production organizations evaluating narrative AI video, previsualization, and repeatable production workflows.
Induce AI Studio includes script upload, scene and shot planning, visual references, generated clips, and final-cut review. Bring a representative scene to evaluate the workflow with your team.
Studio includes shot retakes and final-cut review. Use a pilot to assess the current direction controls and review revised clips in the context of your sequence.
Induce AI’s Terms say customers retain ownership of outputs from their inputs and have commercial usage rights, subject to the Terms and applicable third-party rights.
Creators
Your own script, and characters that stay the same.
Indie filmmakers, solo creators, and small crews who write their own scripts and want to turn them into finished short films without a full production team.
Yes. Paste or upload a screenplay and Induce AI breaks it into scenes and shots you can plan, generate, and review. You can also start from a short brief and develop the script inside the editor.
Induce AI keeps references for your characters, wardrobe, and locations and carries them from shot to shot. The Pro plan adds the Rhapsody engine, which checks each take against the story before you build on it.
Yes. Give a note on a single shot, generate a retake, and review it in the context of the cut. The shots around it stay as they are.
Induce AI’s Terms say customers retain ownership of outputs from their inputs and have commercial usage rights, subject to the Terms and applicable third-party rights.
Starter gives one creator seat with access to all top models. Pro adds the Rhapsody engine and a team workspace for up to five people. See the pricing page for current credits and prices.
Agencies
Agency production, shot refinements and custom setups.
Induce AI brings the script-to-video workflow into a project: develop the brief or screenplay, plan scenes and shots, create visual references and clips, then assemble a cut for review. Agency networks can scope a pilot around their production volume and review process.
Teams can review generated clips and use shot direction or retakes to refine a cut in Studio. Test the revision workflow with your own brief during a pilot; turnaround depends on the production and generation requirements.
Yes. Share your identity, cloud, integration, or export requirements. SSO, VPC deployment, and custom formats may need development. Our team will confirm availability, feasibility, and timing before you commit.
Enterprise
Screenplay to video at volume, SSO and private deployment.
Induce AI is an AI filmmaking workspace for teams developing films from briefs or screenplays. The Story workflow brings scene plans, visual references, generated clips, and a reviewable final cut into one project.
The Story workspace includes script upload. Teams can develop the material through a scene and shot plan, create visual references and clips, and assemble a cut for review.
Induce AI is designed around projects, scenes, shots, and cuts rather than single prompts. Discuss the length, throughput, review process, and support you need with sales before an enterprise rollout.
Discuss the production tools and events you need to connect. Request current API documentation and availability, then agree the scope and timing of any custom integration during your pilot.
Yes. Share your identity provider, cloud, or delivery requirements with our team. SSO, private deployments, and custom formats are scoped requests that may need development. Availability, feasibility, and delivery timing are confirmed before you commit.
Induce AI’s Terms say project content is not used to train Induce AI or third-party models without separate, explicit written consent. Review the Terms and request current security documentation during procurement.
Discuss a pilot with sales using your own script, expected production volume, review steps, and security requirements. Enterprise customers can request current security materials.
Guides and tools
AI dubbing
Dubbing a video into 17 languages in each speaker’s own voice.
AI video dubbing replaces the speech in a video with speech in another language. In Induce AI Remake, the speakers are cloned so they talk in the new language with their own voices.
Yes. Remake clones each speaker's voice and uses it for the new language. Casting a different voice from a library is not available yet.
Lip sync is an optional switch. Without it, only the audio changes and mouths keep the original shapes. With it, a separate pass matches mouth movement to the new language and adds to the cost.
Up to two minutes per upload, as one MP4 or MOV file. Vertical 9:16 video works best.
Seventeen: Hindi, Tamil, Telugu, Bengali, Marathi, Kannada, Malayalam, English, Spanish, Portuguese, French, German, Japanese, Korean, Mandarin Chinese, Arabic and Indonesian.
AI product photography
Product photos by category, and what to check before you publish.
AI product photography means generating product images and videos from photos of the real item, instead of staging a full studio shoot. You supply reference photos, choose the setting, light and whether a model wears or holds the product, and the tool generates the shots. The result still needs checking against the real product before it goes on sale.
You need 1 to 8 clear photos of the same product, each at least 512 px on the shortest side, in PNG, JPG, WebP or HEIC. Include the front, the back or side, and close-ups of logos, labels and any detail a buyer would look for. Plain, even light in your own photos helps the output keep colours right.
It can replace much of the routine work, such as packshots, lifestyle scenes and on-model shots for ads and listings. It is weaker where exact detail matters, such as small printed text, engraving, or colour that must match a grade or a swatch. For those, keep a real photo or check every AI image against the item.
No. Induce AI is credits-based and has no free tier. Flash Shoot shows an estimated credit total for the shoot before you generate anything.
Often yes, but each marketplace sets its own image rules, and they change. Many ask for a plain white background on the main image and an image that shows the actual product being sold. Read the current rules for your marketplace and category, and make sure every image matches the real item.
AI video templates
Product ads, reels, transformations and effects.
An AI video template is a saved recipe for a particular shot: the scene, the camera move and the kind of change that happens. You supply the subject, such as a product photo, and the template handles the rest of the setup so you do not start from a blank prompt.
Open Templates in the Induce AI app, choose a template, add your image or idea and a short direction, then generate. You review the result, generate again if needed and download the version you want.
No. Induce AI has no free tier. Each generation uses credits from your plan, and the pricing page shows the current plans and costs.
Yes, within limits. The direction you write steers details such as setting, colour, mood and pacing. The core move stays the same, so pick the template whose motion is closest to what you want.
Start with the product ad templates: Giant Product for launches, Drink Flyby Bump for drinks, Blue Hour Speaker Film for tech and Color Pop Lookbook for clothing. Transformations such as Liquid Melt also work as a scroll-stopping opener.
Free shot list generator
The free tool that turns a scene into a shot list.
Yes. It is free, needs no account and has no usage limit. It runs in your browser, so there is nothing to install.
No. The shot list is worked out on your own device and nothing you paste is sent to a server or stored. Close the tab and it is gone.
Standard screenplay format works best: scene headings such as INT. KITCHEN - NIGHT, character names in capitals above their dialogue, and parentheticals in brackets. Fountain files pasted as text work too. Plain prose also works; each sentence is read as a beat.
No. It applies standard coverage rules: an establishing wide for each scene, a master and singles for dialogue, and inserts for objects that change. That makes the result predictable. Treat it as a first pass and edit it like any shot list.
This tool does not draw images. Each shot comes with a ready-to-use AI video prompt you can paste into any generator. Induce AI goes further: it reads the script, plans the shots and renders a still for every shot with the same characters throughout.
With the old rule that one screenplay page plays for about a minute, at roughly 180 words a page. It is a rough guide; action-heavy pages usually play longer than dialogue pages.