Alternative to KeywordStock

If you are researching a KeywordStock alternative — or evaluating KeywordStock for the first time — this guide will help you decide. KeywordStock is a Windows and macOS desktop app that reads photos and clips on your own computer, then writes titles, descriptions, and keywords shaped for iStock, Shutterstock, Adobe Stock, and eight more agencies.

Keeping pixels on your PC is a real preference. As of October 2026, KeywordStock still ships a pre-1.0 unsigned installer, does not publish paid credit prices, and markets Getty vocabulary sizes that conflict on the same homepage. This guide covers what the product offers (verified from keywordstock.com/en/, the help page, the download page, the iStock / Getty guide, signup, and terms), who it fits, and the workflow tradeoffs before you commit.

What KeywordStock offers

The public site positions KeywordStock as a local-vision desktop tool for microstock contributors. The website is for signing up and managing credits; the actual work happens in the app. From the English homepage, help FAQ, and download notes as of October 2026:

• On-device reading; only the words travel. The homepage line is explicit: 0 images leave your computer — only the words are sent. Vision runs on your machine. A few kilobytes of generated text go to their servers to be matched against agency vocabulary. Originals are not stored in their cloud, per that marketing.
• Titles, descriptions, and agency-shaped keywords. JPEG, PNG, TIFF, vector files, and MOV/MP4 clips are in scope, plus AI-generated stills in one pass. Declaring work as AI-generated still happens in each agency's own portal.
• Eleven agencies in one run, including PIXTA. The help page lists iStock / Getty Images, Shutterstock, Adobe Stock, Alamy, Dreamstime, Depositphotos, 123RF, Pond5, Freepik, Vecteezy, and PIXTA. Tick the agencies you submit to; metadata is embedded in the file or exported as CSV in the results folder (README.txt documents each agency's steps).
• Getty / iStock controlled-vocabulary matching. The homepage markets both 37,000+ and 39,000+ Getty/iStock terms on the same page — treat that spread as unverified marketing, not a measured dictionary size. Their iStock / Getty guide says keywords must match Getty's vocabulary exactly (including qualifiers such as Landscape - Scenery), up to 50 keywords per file, and that ESP uses Description rather than Title. ESP itself has no CSV upload: embed IPTC/XMP and drag the file in.
• Review that follows your edits. Drag to reorder, click × to drop a term, or promote a suggestion. Help copy says correct 3–5 files and the rest of the batch follows; writing style is remembered for next time.
• Video via multi-moment sampling. MOV/MP4 clips cost 3 credits versus 1 credit per image. Results include paste_helper.html so you can copy fields agency-by-agency, or use that agency's CSV.
• No BYOK key; pay per file. You do not bring your own OpenAI (or other) API key. Credits are charged per file; public pages market no monthly fee. Signup gives 200 trial image credits with no card.
• Thai and English UI. The app and website switch between Thai and English; the default locale at keywordstock.com is Thai. Keywords themselves are written in English, as agencies require.

Public pages document a signed-in desktop workflow. They do not publish a Customer API, a Lightroom plugin, or a browser batch studio.

Who KeywordStock is built for

It earns a fair look in a few specific contributor scenarios:

1. Getty / iStock contributors who refuse to upload originals to a web keyworder. If keeping pixels on the PC is a hard requirement, local vision plus ESP embed is the product's clearest pitch.
2. Desktop-native photographers with 16 GB of RAM (12 GB minimum). The download page states the image reader will not start on 8 GB. This is not a Chromebook or 8 GB laptop tool.
3. Thai/English bilingual contributors. UI language is switchable; Thai is the site's default locale. Generated keywords remain English.
4. Multi-agency uploaders who want PIXTA in the same run. Eleven agencies, including PIXTA (IPTC on JPEG), is broader than many photo-only CSV tools.
5. Contributors who prefer a 200-credit sniff-test with no card. That trial is larger than most keywording signups. It is still a trial, not a published volume rate.

KeywordStock pricing (October 2026)

Public pages as of October 2026 advertise a free trial and a per-file credit model. They do not list paid pack dollars, a subscription ladder, or a $/image rate. The table below is everything we could verify from the homepage, help, signup, and terms — confirm live, and ask for an invoice before you plan a 10,000-file sprint.

Terms (last updated 2026-09-12) say purchased credits are not refundable in cash; failures on their side are refunded in credits. They may change the service or its pricing with notice on the website. Until a SKU is on the site, contributors should treat paid cost as unpublished.

Workflow strengths

Local vision is a real privacy win. Most browser keywording studios upload at least a preview. Here the pitch is stronger: the model reads the file on your disk, and only generated text leaves the machine. For client work, unpublished shoots, or contributors who simply do not want originals on someone else's S3 bucket, that is a legitimate reason to prefer a desktop reader.

Getty-shaped terms, then embed into the file. The homepage demo maps generic vision words (hard hat, safety vest) onto Getty strings (Hardhat, Reflective Clothing). For iStock, their own guide says drop the file into ESP and the fields fill themselves if embedded keywords match the vocabulary. That is a tighter ESP loop than copy-pasting a generic AI list into red chips.

Eleven agencies, including PIXTA, in one tick-list. You are not re-running vision per marketplace. CSV for agencies that accept it sits next to the files; agencies without CSV (Getty ESP, Alamy per their agency rules page) are called out as embed or portal workflows.

A 200-credit, no-card trial. Two hundred files is enough to keyword a real shoot, not a five-image teaser. You can judge title style and Getty matching before anyone asks for a card.

Practitioner framing. The site says it is built by a working Getty/iStock contributor; sample images are from the maker's own factory shoot. That is not a lab-only pitch.

Thai/English UI without a developer key. No BYOK hygiene, no leaked provider key on your bill. Credits are the meter.

Limitations to know before you commit

Paid prices are unpublished. Homepage, help, and signup show 200 trial credits and “pay per file.” There is no public pack ladder and no dollar figure. You cannot compare a 10,000-image month against a published Scale rate until they send an invoice. Terms also allow pricing changes with website notice.

The installer is not code-signed, and the app is still 0.9.x. The download page (last updated 2026-09-28) ships version 0.9.3 for Windows (91 MB), Apple silicon (96 MB), and Intel Mac (89 MB). Windows SmartScreen and macOS Gatekeeper warn that the developer is unknown on every new release; help tells you to click More info → Run anyway. The 0.9.3 notes still fix an “AI image reader not responding” case on vertical 9:16 clips and a stale-connection bug if you leave the app open. That is a pre-1.0 desktop, not a signed studio SLA.

Hardware floor: 16 GB recommended; 8 GB will not start. Download requirements: 16 GB RAM or more, 12 GB minimum, and the image reader will not start on 8 GB. Plan on 10 GB of disk (about 8 GB on first run to download the local model). Internet is required on first launch and while working, because words are sent for vocabulary matching even though images stay put. Windows 10/11 64-bit; macOS 14+ recommended (macOS 13 falls back to CPU). A Vulkan-capable GPU is faster; CPU-only works.

Getty ESP is embed-only — their own guide says there is no CSV upload. Multi-agency CSV exists in the results folder for agencies that take it. iStock / Getty contributors still embed IPTC/XMP and drag files into ESP. If your production line is a 50-keyword Getty CSV into a bulk importer, this is a different motion.

Video delivery is a paste helper, not a footage pipeline. Clips cost 3× stills. Help tells you to open paste_helper.html in the results folder and copy field by field, or use that agency's CSV. Fine for a handful of MOV files; not a dedicated stock-video exporter.

The 37k vs 39k Getty claim is unverified marketing. The same English homepage says 37,000+ terms in the hero and 39,000+ in the stats row. No independent pass-rate is published. Treat the controlled-vocabulary story as a workflow claim (they match strings before embed), not as a proven dictionary census.

No public API or browser job. You install an unsigned desktop binary, sign in with the website password, and work locally. There is no documented Customer API for Photo Mechanic, DAM, or scripted pipelines.

FAQ about KeywordStock

How much does KeywordStock cost?

As of October 2026, signup is free and includes 200 trial image credits with no card (keywordstock.com/en/signup). Help says an image costs 1 credit and a clip costs 3. Paid pack prices are not published on the homepage, help, or signup pages. Terms say purchased credits are not refundable in cash. Ask for an invoice before you budget a large batch.

Are original files uploaded?

Public help says no: the AI reads images on your computer; only the words it produces (a few KB) travel to their servers to be matched against agency vocabulary. Internet is still required while you work, and first launch downloads a local model.

Which agencies are supported?

Eleven, per the help FAQ: iStock / Getty Images, Shutterstock, Adobe Stock, Alamy, Dreamstime, Depositphotos, 123RF, Pond5, Freepik, Vecteezy, and PIXTA. Agency field limits are summarized on the agency rules page.

How do iStock and Getty uploads work?

Their iStock / Getty guide is explicit: there is no CSV upload to ESP. Embed IPTC/XMP, drag the file in, and keywords that match the controlled vocabulary should show green. Up to 50 keywords; homonyms need the full Getty qualifier string. Date and country are separate ESP fields you still fill.

Does it keyword video?

Yes — MOV/MP4. The app samples several moments from the clip. A clip costs 3 credits. Delivery is embed, CSV, or paste_helper.html copy-paste from the results folder.

Why do Windows and macOS warn on install?

The download and help pages say the app is not code-signed yet. Windows shows SmartScreen (“More info” → “Run anyway”). On Mac, cancel the first dialog, then System Settings → Privacy & Security → Open Anyway. The warning returns with every new release. That is their documented behavior, not a third-party rumor.

Looking for an alternative to KeywordStock?

The local-desktop bet is a genuine workflow tradeoff, not a SaaS clone with a different logo. You optimize for pixels that never leave the PC + Getty embed + a Thai/English unsigned 0.9 app. Cloud export-first studios optimize for published volume pricing, 10,000-image jobs, signed browser workflows, and named agency CSVs. For photographers and AI-image sellers who treat metadata as a production line and want that second model, AI Keywording is worth a direct comparison.

AI Keywording is a managed browser-first keywording studio. No desktop install — inference is included via managed tokens, pricing is a public ladder, and agency exports are built in:

Stay with the desktop app if originals must never leave the machine, you already work in Getty ESP via embedded IPTC, you have 16 GB of RAM, Thai/English UI matters, and a 200-credit trial is enough to prove the titles — and you are willing to install an unsigned 0.9.3 build and invoice paid credits privately.

Consider AI Keywording if you need a published volume rate (Growth $59 / 12,000; Scale $129 / 42,000 ≈ $0.003), 10,000-image jobs without a 16 GB RAM floor, named Adobe V2 / Shutterstock / iStock-Getty / Bridge / IPTC-XMP exports including a Getty 50-keyword CSV, output in 100+ languages, a Customer API, PAYG at $0.04/image, or a signed browser workflow instead of Run anyway.

The bottom line

KeywordStock is a serious close-tier option for contributors who value on-device vision, Getty-shaped embed, eleven agencies including PIXTA, and a 200-credit no-card trial. Keeping pixels on your PC is a real preference, and the local reader is the honest reason to choose it.

The core tradeoff is workflow philosophy, not a feature checklist clone: the desktop app optimizes for privacy-first local files and ESP drag-and-drop; export-first studios optimize for published Scale economics, 10K cloud jobs, and agency CSVs you download. Pick the model that matches how you actually ship metadata each month. AI Keywording is one managed cloud alternative when unpublished paid credits, an unsigned pre-1.0 installer, the RAM floor, Getty embed-only, or a video paste helper no longer fit your volume or agency mix.