The AI Museum
AI Analytics1,072 recordsRecounted 23 September 2026

Closed-weight multimodal models from United States of America

44

Models

11 organisations

0%

Open weights

0 open, 44 closed

13

Frontier

By Epoch's flag

7B

Largest traceable

Apollo 7B

2021–2026

Span

7 with a traceable size

Records

44
Closed-weight multimodal models from United States of America, newest first. Each row links to the source Epoch AI records for it.
ModelOrganisationReleasedParametersWeights
GPT-6 AstraOpenAI2026-09Closed
GPT-5.5 InstantOpenAI2026-05Closed
GPT-5.5OpenAI2026-04Closed
GPT-5.5 ProOpenAI2026-04Closed
Muse SparkMeta AI2026-04Closed
GPT-5.4OpenAI2026-03Closed
GPT-5.4 ProOpenAI2026-03Closed
GPT-5.2 ProOpenAI2025-12Closed
Gemini 3 ProGoogle DeepMind2025-11Closed
GPT-5.1OpenAI2025-11Closed
GPT-5.1 InstantOpenAI2025-11Closed
GPT-5 ProOpenAI2025-10Closed
GPT-5OpenAI2025-08Closed
GPT-5 miniOpenAI2025-08Closed
GPT-5 nanoOpenAI2025-08Closed
Llama 4 Behemoth (preview)Meta AI2025-042.0T *Closed
GPT-4o (Mar 2025)OpenAI2025-03Closed
GPT-4o (Jan 2025)OpenAI2025-01Closed
Amazon Nova ProAmazon2024-12Closed
Apollo 7BMeta AI2024-127BClosed
Fugatto 1NVIDIA2024-113BClosed
GPT-4o (Nov 2024)OpenAI2024-11Closed
GPT-4o (Aug 2024)OpenAI2024-08Closed
Claude 3.5 SonnetAnthropic2024-06Closed
GPT-4oOpenAI2024-05Closed
GPT-4 Turbo (Apr 2024)OpenAI2024-04Closed
Reka CoreReka AI2024-0467B *Closed
Claude 3 OpusAnthropic2024-03Closed
Claude 3 SonnetAnthropic2024-03Closed
MM1-30BApple2024-0330B *Closed
Gemini 1.0 ProGoogle DeepMind2023-12Closed
Gemini 1.0 UltraGoogle DeepMind2023-12Closed
Gemini Nano-1Google DeepMind2023-122BClosed
Gemini Nano-2Google DeepMind2023-123BClosed
GPT-4 Turbo (Nov 2023)OpenAI2023-11Closed
OmniVecTensorTour2023-11Closed
PaLI-3Google DeepMind2023-105BClosed
GPT-4VOpenAI2023-09Closed
GPT-4 (Jun 2023)OpenAI2023-061.8T *Closed
PaLI-XGoogle Research2023-0555B *Closed
GPT-4 (Mar 2023)OpenAI2023-031.8T *Closed
MetaLMMicrosoft Research2022-06Closed
NÜWAMicrosoft Research2021-11870MClosed
ALIGNGoogle Research2021-06820MClosed

An asterisk marks a parameter figure the source does not consider traceable to the people who built the model. That judgement is the source’s and it has false negatives: Kimi K3 is marked speculative at 2.8T while Moonshot’s own model card states the figure. Why that distinction carries the whole exhibit.

Counts are of models Epoch AI records as notable, not of every model released. The most recent month is always incomplete, because a model enters the dataset when it is reviewed rather than when it ships.

Epoch AI, 'Data on AI Models'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/data/ai-models-documentation' Used under CC BY 4.0.

How this museum is made