Multimodal models in 2025
19
Models
9 organisations
26%
Open weights
5 open, 14 closed
3
Frontier
By Epoch's flag
424B
Largest traceable
ERNIE-4.5-VL-424B-A47B (文心大模型4.5)
2025
Span
4 with a traceable size
Releases
by monthRecords
19| Model | Organisation | Released | Parameters | Weights |
|---|---|---|---|---|
| GPT-5.2 Pro | OpenAI | 2025-12 | — | Closed |
| HyperCLOVA X SEED 32B Think | NAVER | 2025-12 | 32B | Open |
| Gemini 3 Pro | Google DeepMind | 2025-11 | — | Closed |
| GPT-5.1 | OpenAI | 2025-11 | — | Closed |
| GPT-5.1 Instant | OpenAI | 2025-11 | — | Closed |
| GPT-5 Pro | OpenAI | 2025-10 | — | Closed |
| Qwen3-Omni-30B-A3B | Alibaba | 2025-09 | 35B | Open |
| GPT-5 | OpenAI | 2025-08 | — | Closed |
| GPT-5 mini | OpenAI | 2025-08 | — | Closed |
| GPT-5 nano | OpenAI | 2025-08 | — | Closed |
| Seed-1.6-Thinking | ByteDance | 2025-06 | 230B | Closed |
| Llama 4 Behemoth (preview) | Meta AI | 2025-04 | 2.0T * | Closed |
| Llama 4 Maverick | Meta AI | 2025-04 | 400B * | Open |
| Llama 4 Scout | Meta AI | 2025-04 | 109B * | Open |
| ERNIE-4.5-VL-424B-A47B (文心大模型4.5) | Baidu | 2025-03 | 424B | Open |
| GPT-4o (Mar 2025) | OpenAI | 2025-03 | — | Closed |
| Mistral OCR | Mistral AI | 2025-03 | — | Closed |
| GPT-4o (Jan 2025) | OpenAI | 2025-01 | — | Closed |
| Kimi k1.5 | Moonshot | 2025-01 | — | Closed |
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.