AI Era CodexCard codex · Full pool
Home › Codex

Codex · Full pool

Everything in the T-CLAW-14 generation base (growing with the dataset): filter by era, type and rarity, with “what this card changed” and an official source for every card. Click ☆ to save a card; filters remember your picks.

Data version = T-CLAW-14 generation base, 106 entries (T-GENE-10 merge, 2026-11-05) · Rarity is editorially assigned and used for filtering and browsing only

Archive completeness 94/96(98%)

Scope: the denominator is the number of cards included in the completeness statistic and is recalculated automatically as the pool grows (currently 96 = model cards of the five formal eras; the 7 prehistory reference cards are excluded, and event cards are listed separately with completion=null). The numerator counts cards with all nine fields present and no questionable spec. Still to complete: GPT-4 and the ERNIE 4.0 launch.

Model cards · 3-axis filter

Category

The category axis is live, driven by filter and count data; the 072 delivery folded in 26 products and 6 infrastructure entries; the technology category fills automatically once the 073 dataset lands; the 18 event cards are listed separately below (completion=null).

Era
Type
Rarity

Show all cards

Event cards · listed separately (18 items, excluded from filters and completeness)

Event cards carry completion=null: they count neither toward completeness nor toward the collection-progress numerator. The denominator is fixed at 49 = 56 − 7 (T-AINEW-075).

EventEraDateWhat it changedScopeDetailsShare
Cohere foundedPrehistory2019A Toronto-born enterprise LLM company built around retrieval augmentation and private deploymentcompletion=nullTimeline ↗
Stability AI foundedPrehistory2019Its open-weights strategy set off the image-generation ecosystem; cumulative model downloads passed 350 millioncompletion=nullTimeline ↗
MiniMax foundedAPI Era2021-12Yan Junjie leaves SenseTime to found the company, among the earliest general-AI entrantscompletion=nullTimeline ↗
ModelBest foundedAPI Era2022-08Spun out of Tsinghua's NLP lab, it established the high-efficiency on-device “small but mighty” routecompletion=nullTimeline ↗
Microsoft invests further in OpenAIMultimodal Era2023-01-23A hyperscaler places its bet as the ChatGPT shockwave is fully capitalizedcompletion=nullTimeline ↗
Baichuan AI foundedMultimodal Era2023-03-24Founded by Wang Xiaochuan with open, commercially usable Chinese base models, joining the domestic front tiercompletion=nullTimeline ↗
Moonshot AI foundedMultimodal Era2023-04Founded by Yang Zhilin, it raised over $1 billion within its first yearcompletion=nullTimeline ↗
Mistral AI foundedMultimodal Era2023-04A European frontier lab founded to break the US closed-model monopoly with open weights and strong price-performancecompletion=nullTimeline ↗
StepFun foundedMultimodal Era2023-04-06Founded by former Microsoft global VP Jiang Daxin, betting on multimodality and scaling lawscompletion=nullTimeline ↗
01.AI founded and the Yi series open-sourcedMultimodal Era2023-05Founded by Kai-Fu Lee, a unicorn within six months, with the open Yi series taking on Llama 2completion=nullTimeline ↗
DeepSeek foundedMultimodal Era2023-07Incubated by quant giant High-Flyer, with Liang Wenfeng leading its entry into large modelscompletion=nullTimeline ↗
Interim Measures for the Management of Generative AI Services publishedMultimodal Era2023-07-13Among the world's earliest dedicated generative AI rules, establishing the filing regimecompletion=nullTimeline ↗
The first batch of large models approved for public accessMultimodal Era2023-08-31Chinese LLMs are cleared for consumers, opening the first year of universal accesscompletion=nullTimeline ↗
Lovable becomes a unicornAgent Reasoning Era2025-07-17Vibe coding reaches the masses and it becomes one of the fastest unicorns evercompletion=nullTimeline ↗
OpenAI restructuring (PBC)Agent Reasoning Era2025-10-28Its capital structure is settled and the Microsoft relationship is repricedcompletion=nullTimeline ↗
The DeepSeek shockwaveOpen-Source Era2025-01-27DeepSeek tops the US App Store as global capital reprices Chinese AIcompletion=nullTimeline ↗
Qwen has the most derivative models in the worldOpen-Source Era2025-02-10Qwen overtakes Llama to become the largest open-model family in the worldcompletion=nullTimeline ↗
MiniMax lists in Hong KongOpen-Source Era2026-01-09The first IPO by a Chinese LLM company, breaking the ice for AI asset securitizationcompletion=nullTimeline ↗
AI Era Codex · Loop site family · Source: T-CLAW-14 generation base, 106 entries · Yunnan ICP filing 2023002020-2