When the AI World's "Household Registry" Gets Reorganized: easy-learn-ai's Model Catalog Restructure
*Source commit: e6c189a*
How would you organize a registry of every AI model in the world — by release date, by country, by capability, or by the company behind them? When you're facing dozens of vendors and hundreds of models spanning text, image, and video, the answer isn't obvious.
The Problem: One Giant "Encyclopedia"
In the easy-learn-ai open-source project, all model information used to live in a single massive JSON file — model.json — over 5,000 lines long, like an encyclopedia with no table of contents. It contained OpenAI's GPT series, Anthropic's Claude series, Google's Gemini series, China's DeepSeek, Alibaba's Qwen, Baidu's ERNIE, ByteDance, plus image generators like Midjourney, DALL·E, and Stable Diffusion, and video generators like Sora and Veo.
It was like having every resident of your apartment complex written on one big sheet of paper — no building number, no unit, no door number. Finding anyone meant reading from top to bottom.
The Fix: Organizing by "Family"
Commit e6c189a did something simple but far-reaching: it split the encyclopedia into 20 vendor-named booklets. US vendors (OpenAI, Anthropic, Google, Meta, xAI, Midjourney, Runway, Pika, Stability AI), Chinese vendors (Alibaba, Baidu, ByteDance, DeepSeek, Moonshot AI, MiniMax, Tencent, Kuaishou, Zhipu AI), and Germany's Black Forest Labs each got their own file.
Why does this matter?
1. Faster lookup. Want to know DeepSeek's models? Just open deepseek.json — R1, V3, V4-Pro, V4-Flash, Math-V2, OCR — with each model's context window, max output length, open-source status, release date, and links.
2. Easier comparison. Open alibaba.json and openai.json side by side to compare Qwen3.7-Max and GPT-5.5.
3. Simpler maintenance. A new Alibaba model? Update one file without touching anyone else's data.
Highlights from the Catalog
DeepSeek: the Open-Source Dark Horse
deepseek.json records 18 models, from DeepSeek LLM (January 2024) to DeepSeek-V4-Pro. Its MoE (Mixture-of-Experts) architecture uses 671B total parameters but activates only 37B per query — like a panel of specialists where only the relevant experts answer. V4-Pro reaches 1.6T total / 49B active parameters with a 1M-token context window.
Alibaba Qwen: the All-Rounder
alibaba.json covers text (Qwen3), vision (Qwen3-VL), code (Qwen3-Coder), image generation (Qwen-Image, Z-Image), video generation (Wan2.2, Wan2.5), multimodal (Qwen3-Omni), plus reasoning models QwQ and QVQ. The open-source Qwen3.5-Plus has 397B total / 17B active parameters, up to 19× faster inference than its predecessor, with API pricing around 1/18 of Gemini 3 Pro.
OpenAI: the Precision Instruments
From GPT-4 to GPT-5.5, from o1 to o4-mini, from DALL·E to Sora. Notably, GPT-5.5 supports up to a 1.05M-token context window, and GPT-5.3-Codex scored 56.8% on SWE-Bench Pro — solving over half of software engineering tasks autonomously.
Google Gemini: the Multimodal Chameleon
From 2.0 Flash to 3.5 Flash, from Nano Banana (image editing) to Imagen 4 (text-to-image), Veo 3 (video) and Gemini 3 Pro (all-purpose reasoning). Gemini 3.5 Flash beats the larger Gemini 3.1 Pro on coding and agent benchmarks like Terminal-Bench 2.1 and MCP Atlas while keeping Flash-series speed.
A Quiet Standardization Movement
Every model entry follows a unified schema:
modelNamecompanycountryopenSourceStatusreleaseDatedescriptionmodelTags(text generation, vision understanding, code enhancement, tool calling, etc.)contextWindowmaxGenerationTokenLengthrelatedLinks
Why It Matters to Everyone
AI models iterate monthly or even weekly. Today GPT-5.5 leads; tomorrow Qwen3.7-Max may surpass it on some task; the day after, DeepSeek releases a cheaper new version. In this fast pace, information itself is a competitive advantage — whoever finds the best-fit model fastest gets the best results at the lowest cost.
easy-learn-ai's work is like building a bridge across a rushing river, so ordinary users can cross safely instead of being swept away by the flood of information.
Looking Ahead
This restructure is just a beginning. The catalog may add pricing, inference speed, energy consumption, multilingual capability scores, safety ratings — even live API integration for one-click real-world comparisons.
As the AI world grows more complex, navigation tools like easy-learn-ai become more important. However powerful a model is, if you can't find it, use it, or choose it, it might as well not exist.
---
*This article is based on commit e6c189a of the easy-learn-ai project, which catalogs global AI model information to help developers and users better understand and choose AI tools.*