The Storyblocks metadata requirements come down to three hard limits: titles between 16 and 140 characters, at least six keywords and no more than 100, and all of it in English. What trips contributors up is rarely the limits themselves. It is that Storyblocks reads metadata from three different places depending on what you upload, and a footage clip can arrive on the site with its title already decided by an IPTC field you never opened.
What you are actually submitting to
Storyblocks runs on subscriptions rather than per-download sales. Members pay monthly or annually and download from the Member Library, and contributors earn through the Member Library Partner Program, where your share depends on how much of your content gets downloaded relative to everyone else's and on revenue from member subscriptions. Storyblocks describes this as an earnings model that can offer consistency, which is a fair way to put it: there is no single big sale, and no single dry month either.
Two things follow from that model. The library is curated rather than open, and Storyblocks states that roughly 5% of applications are accepted. It also operates a strict zero generative AI policy, describing its library as 100% human-made in a company statement updated in May 2026. If you have been building an AI-generated library, no metadata field will get it in here, and the agency-by-agency rules for AI content are worth checking before you plan a submission run.
As of now Storyblocks takes applications for HD, 4K and VR/360 equirectangular footage; animations, including motion graphics, backgrounds, alpha channel effects, transitions and lower thirds; templates for After Effects, Premiere Pro in both PRPROJ and MOGRT form, Apple Motion and DaVinci Resolve; and exclusive music in any genre. The metadata rules below are shared across those types, with one exception for audio that comes up later.
The title field: 16 to 140 characters
Storyblocks states a minimum title length of 16 characters and a maximum of 140, and suggests aiming for about 65 including spaces. That target is the useful number. It is long enough for a full descriptive sentence and short enough that nothing gets cut off in a search result, which is the shape Storyblocks wants: a title is a sentence describing the clip, not a list of keywords stacked up with commas.
The structure it asks for is who, what and where. The subject should appear along with what is happening, where it is happening, or both. Storyblocks' own example is "Father throws baseball to son in living room" — subject first, action, then setting. Mood belongs in the title when it is genuinely part of the clip, and so does time of day or season when either is obvious on screen.
Beyond that core, Storyblocks suggests adding shot type or scene view where it helps a buyer: rack focus, aerial or drone, close-up, back or front view. Notable clothing and props earn a place too, since they are often exactly what a buyer is searching for. The example it gives is "Doctor with stethoscope and lab coat examines patient", where the props do more work than any adjective would.
Three rules sit alongside the length limit, and all three reject in practice. The title must not match the filename, so an exported clip name pasted into the title field will come back. Every file needs its own title even when clips are near-identical, and Storyblocks is explicit that you must change at least one word for the title to be accepted — a real problem if you are submitting a 40-clip sequence from one shoot. And trademarked names are not allowed, including brand names that have drifted into generic use.
Title length limits across three agencies
Storyblocks accepts titles of 16 to 140 characters and suggests about 65, while Adobe Stock recommends staying under 70 against a field that allows 200, and Shutterstock permits up to 2,048 characters.
The keyword field: six minimum, 100 maximum
Six keywords is the floor and 100 is the ceiling. Between those, Storyblocks recommends 15 to 30 in its metadata requirements, while its footage keywording guidance says to aim for at least 15 and notes that up to 50 can contribute to a clip's success. Those two ranges are not in conflict so much as differently cautious: 15 is the point below which a clip is probably under-described, and somewhere past 50 you are adding words that do not describe the clip.
The unit Storyblocks asks for is a single-concept search term. Single words are fine, and so are short connected phrases where the phrase is the concept — "New York" and "computer screen" are its examples. What it does not want is a sentence fragment pretending to be a keyword, or the same idea split across four near-synonyms.
For what to include, the guidance is close to how a buyer actually searches. Take the words from your title and enter them individually. Add related nouns, verbs and adjectives. Cover the subject, the setting — time of day, season or location — and the action. Then add connected concepts, pairing an emotion with an event or combining two related ideas: a birthday with fun, vegetables with health. Storyblocks also points contributors at relevant terms from its in-demand dashboard, which is the one input on this list you cannot guess from the clip itself.
The things to avoid are short and worth taking literally. Unrelated words are described as hurting the clip's placement and earnings, which is a stronger statement than most agencies make — padding is not neutral here, it costs you. Unnecessary repetition of words, and spelling errors in either field, are the other two. Storyblocks ranks keywording second only to clip quality in what determines whether a file finds its buyer, and the general principles of keyword selection apply here as much as anywhere.
Everything has to be in English
Storyblocks is unambiguous on this: all titles and keywords must be in English. There is no translation layer and no second-language keyword field. If your workflow keywords in your own language first — which is a sensible way to work, because you describe a scene more precisely in the language you think in — the Storyblocks branch of your pipeline needs a translation pass before delivery, not after. This is one of the places where keeping one canonical metadata set and deriving each agency's version stops being a nicety.
Three routes metadata takes into Storyblocks
This is the part of the Storyblocks setup that is genuinely different from other agencies, and the part most worth understanding before a large batch. Metadata can arrive three ways, and which one wins is not always obvious.
How metadata reaches a Storyblocks submission
Storyblocks takes metadata three ways: typed on the platform under the Content tab, attached as a CSV for footage, animations and templates, or embedded as IPTC in a video file, where the Headline field fills the site's Title and falls back to Description and then Title.
The IPTC fallback order is the quiet failure mode. Plenty of editing and cataloguing setups write a short internal note into Headline and the real descriptive sentence into Description. On Storyblocks the Headline wins, so an internal note becomes your public title and your carefully written sentence never appears. If you are not sure what your files currently carry, it is worth reading what each embedded metadata field actually does before a batch goes out. The CSV route has the opposite character — it is unusually generous. If you already build a Shutterstock CSV, that same file is accepted here as-is, which removes an entire reformatting step from a multi-agency delivery.
What review actually sends back
Storyblocks says it may ask you to edit and resubmit metadata that does not meet its requirements. That is a gentler outcome than a hard rejection, but it is the expensive one at volume: a round trip on a batch of two hundred clips is a day you did not plan for, and it arrives after you have already moved on to the next shoot.
The checks worth running before upload rather than after are all mechanical, which means a script or a tool can do them: every title between 16 and 140 characters, every title different from its own filename, no two titles identical across near-duplicate clips, at least six keywords on every file and nothing over 100, no trademarked names in either field, no spelling errors, and the whole set in English. None of that needs judgement. All of it is a resubmit if it slips.
Fitting Storyblocks into a multi-agency pipeline
Storyblocks is strict where other agencies are relaxed and relaxed where others are strict. English-only metadata, no AI-generated content and a title that must differ from the filename are all tighter than the norm. Accepting a competitor's CSV format without edits is considerably looser than the norm. If you also deliver to Pond5, that CSV allowance means one file can serve two agencies, and it is worth knowing where Storyblocks sits among the footage platforms open to contributors before you decide how much of your library to point at it.
The pattern that survives contact with four or five agencies is the same one every time: write the metadata once, at the level of the clip, then derive each agency's version from it. Keywording the same footage three times in three portals is not just slow, it produces three slightly different descriptions of one clip, and the differences are never deliberate.
That is the shape Rastock AI is built around: generate the title, description and keywords once, hold them against each agency's own field rules and banned terms before anything leaves your machine, then deliver over FTP or SFTP to 10+ agencies with the per-agency differences applied rather than hand-edited. There is no revenue share and no ownership claim on your files, and IPTC, CSV and XML export stay yours to take elsewhere. What it does and what it costs are both set out plainly, and the 14-day trial does not ask for a card.
Frequently asked questions
What are the Storyblocks metadata requirements?
Every file needs a title between 16 and 140 characters, with about 65 suggested, and at least six keywords up to a maximum of 100. Titles and keywords must be in English, the title must not match the filename, and neither field may contain trademarked names. Storyblocks may ask you to edit and resubmit metadata that misses any of these.
How many keywords does Storyblocks require?
Six is the minimum and 100 is the maximum. Storyblocks recommends 15 to 30 in its metadata requirements, and its footage keywording guidance suggests at least 15, noting that up to 50 can contribute to a clip's success. Past that, relevance beats volume: unrelated keywords are described as hurting a clip's placement and earnings.
Can I submit a Shutterstock CSV to Storyblocks?
Yes. Storyblocks accepts CSV files formatted for Shutterstock or Pond5 without editing, alongside its own downloadable template. The CSV route covers footage, animations and templates. Audio artists cannot attach a CSV at this time and need to enter metadata on the platform under the Content tab instead.
Does Storyblocks accept AI-generated content?
No. Storyblocks operates a strict zero generative AI policy and describes its library as entirely human-made, in a company statement updated in May 2026. Metadata cannot work around this, so an AI-generated library needs a different destination. Storyblocks also states that roughly 5% of contributor applications are accepted.
Why did my Storyblocks title come out different from what I wrote?
For video, Storyblocks reads embedded IPTC fields and fills the site's Title from Headline. If Headline is empty it falls back to Description, then to Title. Many editing setups put a short internal note in Headline and the real sentence in Description, so the internal note wins and the intended title never appears.
Do Storyblocks titles and keywords have to be in English?
Yes. Storyblocks states that all titles and keywords must be in English, with no second-language field and no translation layer. If you keyword in another language first, the translation has to happen before delivery rather than after, which is one reason to hold a single canonical metadata set per clip and derive each agency's version from it.