# The catalog says 300 tools and 500 tools in the same file > AI-Catalog is a Shell-and-Markdown list of AI tools with a legend of emoji markers for price and freshness, curated by one person browsing the web. Its header claims 300-plus tools across 45-plus categories, its getting-started section claims 500-plus tools across 30-plus, and it promises every entry is tested and verified one paragraph before admitting updates depend on what the author happens to see. **mehmetkahya0/AI-Catalog** — Huge AI models catalog. A curated list of AI tools, platforms, and resources across various domains. - Repository: https://github.com/mehmetkahya0/AI-Catalog - Website: https://mehmetkahya0.github.io/AI-Catalog/ - Stars: 491 · Forks: 85 - Language: Shell - License: GPL-3.0 - Published: 2026-09-14 · Updated: 2026-09-14 · Language: en - Canonical page: https://hysenlabs.com/projects/mehmetkahya0-ai-catalog ## Two sizes, two category counts, one file The header describes a curated collection of more than 300 AI tools across more than 45 categories, and the quick stats block repeats 300-plus tools and 45-plus categories. Twenty lines later the getting-started section says the repository contains more than 500 carefully curated AI tools across more than 30 categories. Neither number can be right alongside the other, and the category count moves in the opposite direction from the tool count, which is the part that suggests a copy-paste from an earlier revision rather than a rounding difference. A reader sizing the catalog for a decision has to pick one, and neither figure comes with a way to check it, since there is no generated index and no per-category count. The table of contents does show the intended shape, with nine top-level areas and named subcategories under each: creative AI split into text to image, video, audio editing and 3D, content and writing, developer tools including low-code and no-code, assistants and chat, education, business and productivity, specialized tools, entertainment, and an experimental section that holds autonomous agents. ## Tested and verified, then described as opportunistic The curation method is described twice as well, and the two descriptions are incompatible. The getting-started paragraph says each tool is tested and verified to ensure quality and relevance. A note near the top says something different: the author lists the AIs seen while browsing the internet, so updating the list can take a while. That is an honest description of a personal list, and it is not what tested and verified means. Nothing in the repository records what a test consisted of, when it happened, or whether entries are revisited, so a marker on a row tells you when the author last saw the product, not when anyone evaluated it. ## An entry is a link, up to three emoji and one sentence The unit of curation is small. Each row is a bold link, a set of markers from a five-item legend, and a single descriptive clause, for example Midjourney v7 marked new and paid with a note that it supports text to video, or RenderFlow AI described as a multi-model platform bundling several named image and video models. The legend covers free, paid, freemium, new and popular, which is the most information any entry carries. Coverage is broad enough that whole categories are one heading deep: specialized tools holds data analysis and business intelligence, SQL tools, Chrome extensions and an AI detection section, while entertainment covers gaming and music. There is no pricing figure, no context window, no API availability line and no date of testing, so the list tells you a product exists and roughly what it costs, which is one step short of the comparison most people want when they open a catalog. ## Freshness claims stop in 2025 Two dates in the document matter for judging it. The centred footer says last updated January 28, 2026, alongside a next goal of 500 stars, and the last push to the default branch was on 2026-03-18, seven weeks after that line was written and roughly six months before now. The newest entries carry parenthetical dates of April 2025 and March 2025, so the most recent product claim in the text describes a version from a year before the last commit. The header also asserts longevity, that this is one of the first AI-list repositories on GitHub with a first commit on Jun 30, 2023, next to a traffic-counter badge from the awesome-list community, which measures page views rather than accuracy. The repository publishes no releases and there is no changelog of entries, so there is no way to see what changed or when. For a list whose value depends on being current, that is the first thing to check before trusting any entry. ## Shared links and aggregator rows Some entries are looser than the format implies. Three FLUX variants, Pro, Ultra and Kontext, all point at the same vendor page, so the list distinguishes versions without telling you where to sign up for each. One aggregator entry is filed under text to image while describing a platform that wraps GPT-Image, Imagen, Flux Pro Ultra, Midjourney, Kling and Veo3, which puts image and video models in the same row. None of that is dishonest; a directory is not a review. It does mean that when you are comparing two rows, sometimes one of them is a store and sometimes both are, and the reader has to work that out from the text. ## The scripts directory and a checklist from awesome.re The repository is more than one Markdown file. Alongside it sit a `scripts/` directory, a contributing guide, a code of conduct, a roadmap, a summary of improvements and `AWESOME_LIST_CHECKLIST.md`, the checklist used by the awesome-list community, which the header links to. The primary language of the repository is recorded as Shell, and the README's footer credits Claude Sonnet 4 with writing the shell scripts, which explains both the language detection and why there is tooling here at all: which suggests checks over links, formatting or the list structure. None of it is documented in the visible text, so a contributor has to read the scripts to find out what they do. Support runs through GitHub Discussions, the issue tracker, a direct email address and LinkedIn, with the star history chart linked from the header as the closest thing the project has to a dashboard. ## Where this catalog is the wrong tool A link list is a map of what exists, and maps go stale, which is why the freshness gap above is the deciding factor rather than a footnote. For a purchase, a security review or an architecture decision, none of these rows is evidence: no entry says what a tool costs at your volume, whether it has an API, whether your data leaves the machine, or whether anyone has run it in production. For orientation, the list is genuinely useful precisely because it is cheap and honest about being personal. The licence adds one more wrinkle worth naming: GPL-3.0 covers the list itself, while each linked product carries its own terms, so re-publishing the catalog inside a product of your own raises questions the repository does not address. The practical replacement for a list like this is not a bigger list but a shortlist you have run yourself: pick the two or three entries that survived the marker legend, read each vendor's own documentation for pricing and data handling, and treat the catalog as the place you found them rather than the reason you chose them. ## Conclusion ai-catalog fits orientation rather than procurement: it is a reasonable place to find the names in a category you have not read about in months, and the markers tell you at a glance whether something is free or paid. It does not fit a buying decision, since no entry carries an evaluation and the counts contradict each other. Verify two things before you rely on it: what the last commit date says about a specific entry, because the newest dates in the text are from 2025 while the last push was on 2026-03-18, and whether the GPL-3.0 licence on the list matters to you, since it covers the list and not the hundreds of third-party tools it points at. ## FAQ ### How many AI tools does AI-Catalog list? The document gives two answers. The header and the quick stats block say more than 300 tools across more than 45 categories, while the getting-started section says more than 500 tools across more than 30 categories. The repository publishes no index that would settle which figure is current. ### How is AI-Catalog curated and updated? The README says each tool is tested and verified for quality, and separately notes that the author lists the AI tools seen while browsing, so updates take a while. Pull requests are welcome, with a contributing guide and an awesome-list checklist in the repository, and the footer asks for stars toward a goal of 500. ### What do the emoji markers in AI-Catalog entries mean? Five markers from the legend: free for completely free tools, paid for a subscription or one-time payment, freemium for a free tier with premium options, new for recently added tools, and popular for highly recommended by the community. ### When was AI-Catalog last updated? The README states last updated January 28, 2026, and the default branch was last pushed on 2026-03-18. The repository publishes no releases, and the newest product dates mentioned inside the entries are from March and April 2025. ## Sources - [Issues](https://github.com/mehmetkahya0/AI-Catalog/issues) - [License: GPL-3.0](https://github.com/mehmetkahya0/AI-Catalog/blob/main/LICENSE) - [mehmetkahya0/AI-Catalog on GitHub](https://github.com/mehmetkahya0/AI-Catalog) - [Project website](https://mehmetkahya0.github.io/AI-Catalog/) - [README](https://github.com/mehmetkahya0/AI-Catalog/blob/main/README.md) --- Hysen Labs editorial analysis, written from the project's own repository and release notes. Cite the canonical page: https://hysenlabs.com/projects/mehmetkahya0-ai-catalog