May 22, 2026
Price tags, but make it drama
Models.dev: open-source database of AI model specs, pricing, and capabilities
AI shoppers just got a giant cheat sheet — and the crowd is already asking for more
TLDR: Models.dev launched as a free, open database to compare AI tools by price and features in one place. Commenters are impressed, but they’re already demanding better filters, speed data, and reliable updates so the list doesn’t become outdated.
A new project called Models.dev is trying to become the master list of AI models — basically a giant open, community-edited cheat sheet showing what each model can do, how much it costs, and who makes it. You can browse it on the site, pull the data from an API, and even help update it yourself through simple text files in a public code repository. In plain English: it wants to be the one-stop menu for the exploding AI model buffet.
And the comment section? Instant approval mixed with classic internet nitpicking. One fan called it an "absolute gem" and bragged that the table is now "fast af," which is about as close as tech people get to throwing confetti. Others were into the idea but immediately turned into unpaid product managers: where are the filters, why is horizontal scrolling hiding details, and can we please get latency numbers too because the listed price is not the whole story? That became the big mini-drama: is cheap AI really cheap if it’s also slow?
Then came the long-term anxiety. Several commenters loved the neat pricing table but worried that a community project like this lives or dies on updates. AI prices change constantly, and one user basically asked for a full price-history tracker so people can watch the costs fall over time. Another popped in with a very internet-specific question about a supposedly free model elsewhere, adding a dash of chaos to an otherwise glowing launch. The vibe was clear: people want this badly — and now they want it bigger, faster, clearer, and impossible to go out of date.
Key Points
- •Models.dev is introduced as an open-source, community-contributed database for AI model specifications, pricing, and capabilities.
- •The project provides API access through `https://models.dev/api.json`, using Model IDs that match AI SDK identifiers.
- •Provider logos are served as SVG files via a provider-specific URL pattern, with a default logo used when needed.
- •Project data is stored in a repository as TOML files by provider and model, and SVG files for logos; this data powers both the site and API.
- •The contribution workflow supports adding providers, logos, and detailed model definitions, including pricing, token limits, modalities, dates, and OpenAI-compatible provider configurations.