AI models
Megrez-3B-Omni
Training compute
3.6×10²² FLOP
Parameters
3B
Published
Jun 12, 2024
Megrez-3B-Omni is an AI model developed by Infinigence AI, Tsinghua University and Shanghai Jiao Tong University (China), first published in June 2024. It works in the multimodal, language, vision and speech domain, on tasks such as text summarization, image captioning, character recognition (ocr) and 4 more.
Training it took an estimated 3.6×10²² FLOP of compute (estimation method: operation counting). The model has 3,000,000,000 parameters. It was trained on roughly 2T datapoints. Training ran on NVIDIA H100 SXM5 80GB.
Access: Open weights (unrestricted). Its weights are openly available. It is built on top of SigLIP 400M. Epoch AI rates the confidence of this record as confident.
Full record
- Organization
- Infinigence AI, Tsinghua University, Shanghai Jiao Tong University
- Country of organization
- China
- Domain
- Multimodal, Language, Vision, Speech
- Task
- Text summarization, Image captioning, Character recognition (OCR), Visual question answering, Language modeling/generation, Question answering, Speech recognition (ASR)
- Training compute
- 3.6×10²² FLOP
- Compute estimation method
- Operation counting
- Parameters
- 3,000,000,000
- Dataset size
- 2T
- Training hardware
- NVIDIA H100 SXM5 80GB
- Model accessibility
- Open weights (unrestricted)
- Open weights
- Yes
- Base model
- SigLIP 400M
- Epoch confidence
- Confident
SourceEpoch AI, 'AI Models'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/ai-models. Licensed under CC BY 4.0.
← All ai models