CardSight AI adds MMA trading cards and tops 8,250 supported sets

CardSight AI said it now fully supports mixed martial arts trading cards, covering more than 250 MMA releases across 30 years and expanding its catalog to more than 8,250 supported sets. The move extends the company’s API-first trading card infrastructure for developers building apps, databases and AI tools.

Why it matters: - CardSight AI is broadening its trading card infrastructure into MMA, a category that is harder to identify than team sports because cards often lack uniforms, logos and other easy visual markers. - The expansion gives developers access to MMA identification, catalog and market data through the same API used for other sports and trading card games. - The release pushes CardSight AI’s supported set count past 8,250, strengthening the platform’s pitch as a single integration for trading card apps.

What happened: - CardSight AI announced full support for MMA trading cards. - The platform now covers more than 250 MMA releases spanning 30 years, including 1990s cards and current UFC-licensed products from publishers such as Topps and Panini. - The same update adds more than 1,500 sets across sports and trading card games. - The company said the new coverage is live now for all customers, including free-tier users.

The details: - CardSight AI says its system can identify a photo taken at an angle with several cards in frame and return the specific card and parallel in under 300 milliseconds per card at 99.5% accuracy. - The platform identifies graded slabs from PSA, BGS, SGC, CGC and TAG, and extracts the grade. - CardSight AI says live-video support can surface a card as soon as it appears on camera during a break or live-selling stream. - Every identified MMA card resolves to an entry in the CardSight AI Catalog. - MMA market data is published as transparent bid and ask information tied to specific sales. - CardSight AI said it does not calculate comps or publish a single price. - The catalog spans baseball, football, basketball, hockey and MMA on the sports side, and Pokémon, Magic: The Gathering and One Piece on the trading card game side. - The database includes more than 13 million cards organized by segment, manufacturer, release, set, card and parallel. - Printed language detection covers eight languages for trading card games. - New releases continue to ship in days, and the company said it supported Pokémon Pitch Black and 2026 Topps Flagship Football within days of release. - MMA products will follow the same cadence as new releases arrive.

Between the lines: - CardSight AI is positioning itself as infrastructure, not a consumer app, in a market where some companies are trying to add APIs after the fact. - The company is leaning hard into openness: developers can create an account, get an API key and start on a free tier without a credit card. - CardSight AI also offers SDKs for Node.js, Python, Swift, Java and .NET, plus MCP endpoints for AI agents and LLM integrations. - Those MCP endpoints already work with AI coding assistants such as Claude Code and Codex, which lowers the lift for builders using AI to wire in trading card features. - Over the past quarter, thousands of users have used the platform to take a trading card app or website from idea to deployment in a single weekend, the company said. - CEO Eric Nusbaum said the company does not screen what customers are building and does not compete with people who build on the platform.

What's next: - CardSight AI said new MMA trading card products will become identifiable as they reach shelves, with no changes needed on the builder side. - Developers can start at CardSight AI, test the API in the browser at Playground and review documentation for the REST API, SDKs and MCP endpoints. - The company will continue expanding its catalog and market-data coverage as new sets release.

The bottom line: - CardSight AI is using MMA to showcase a broader strategy: one developer API for identifying cards, organizing data and powering trading-card apps across sports and games.