Know whether your game concept has a market before you spend a year building it. Paste a concept or upload your GDD, pick your target platform, and Gameloom scores it against real shipped games in the same genre and platform cohort, inside the Editor.
WHAT YOU GET
- Score across five pillars: Market Opportunity, Core Concept, Economy, Execution and Scalability. With it, a "Why this score" verdict and a "What to improve" action plan, split into Critical and Important fixes.
- A verdict on how your concept sits against its cohort, with the confidence level and the number of similar games behind it.
- Your brief, extracted. Title, genre, platform, pitch, core loop and hook, pulled out of your own text. Good for spotting what your GDD leaves unsaid.
- Your whole document, read. PDF and DOCX up to 50 MB, text and images: charts, tables, diagrams, roadmaps and mock-ups. The first 260,000 characters are analyzed, and the summary tells you what was read.
- Comparables with box art, release year, platforms, review counts, wishlists and price. Each one links to its store page, so you can check any number yourself.
- The ones that died, separated from the ones that survived, plus the failure pattern they had in common.
- Risk cards. A named failure mode, a severity, why it applies to your concept, and what to do about it.
- Price band and median positive-review rate for your genre on your platform.
- Player sentiment across the cohort: what reviewers keep praising, and what they keep complaining about.
- Suggested next actions, based on what went wrong for the games that failed.
- Markdown export. Copy the report to your clipboard, or save it as a .md file into your design docs.
HOW IT WORKS
- Open Window > Gameloom.
- Paste your concept, or upload a GDD as .txt, .md, .pdf or .docx (PDF and DOCX up to 50 MB; Gameloom reads their text and their images, such as charts, tables and diagrams). Text files can be added several at a time; choose which to analyze. A single paragraph works. A full GDD works better.
- Pick your target platform: iOS, Android, Steam, Nintendo Switch, PlayStation or Xbox. If you have stated a platform in your text, Gameloom reads it and preselects it.
- Click Analyze. Gameloom reads your GDD once, tells you exactly what it read, and opens the Score tab. A large PDF or DOCX takes 1 to 3 minutes to read. Score, Analysis and Comparables then run on the extracted brief without re-reading the document. Clear starts over.
WHAT IT DOES NOT DO
Gameloom scores your design, meaning your concept or GDD: its text, and the images in a PDF or DOCX you upload. It does not read, analyze or modify your scene, scripts, prefabs or assets. It adds no runtime code, and nothing from this package ships in your player builds.
REQUIREMENTS
- Unity 2022.3 or newer.
- Dependency: com.unity.nuget.newtonsoft-json (MIT licence), installed automatically by the Package Manager. Nothing to configure.
- An active internet connection. This package is a client for Gameloom's hosted service and does not function offline.
- No account, no sign-in, no API key and no Unity AI credits are required. There is no subscription and no in-package purchase. A light per-IP throttle protects the free service from abuse, at roughly ten analyses a minute. Normal use will not reach it.
PRIVACY AND DATA
- Only the concept or GDD you explicitly choose to analyze (the text, or the PDF/DOCX file you upload), plus the platform you select, are sent to Gameloom's servers.
- No Unity project data is read or transmitted: no scenes, scripts, prefabs or assets.
- This package runs anonymously. There is no sign-in mode and no account to create, so nothing is tied to a user identity. There is no per-install or device identifier.
- No analytics, advertising or tracking SDKs are included.
- Processing: Supabase (hosting and the temporary file store) and Vercel for infrastructure. The extraction step, including reading the pages or pictures of an uploaded file, uses a third-party AI model accessed through OpenRouter (currently an Anthropic model). Market comparison, scoring, risk and benchmark functions use no AI model and run on Gameloom's own infrastructure.
- Your submitted content is not used to train AI models, is not sold, and is not stored in Gameloom's application database.
- Uploaded PDF/DOCX files are stored temporarily in a private store only to read them, deleted right after reading and in all cases within 1 hour.
- Full policy: gameloom.ai/privacy
SUPPORT
- Documentation: gameloom.ai/docs/unity
- Questions, bugs and data requests: contact@gameloom.ai
- Published by gameloom.ai Ltd, registered in England and Wales (company no. 16545032).





