SigLIP 2

Training compute

8.2×10²² FLOP

Parameters

1.1B

Published

Feb 20, 2025

SigLIP 2 is an AI model developed by Google DeepMind (United States), first published in February 2025. It works in the vision domain, on tasks such as image classification and image embedding.

Training it took an estimated 8.2×10²² FLOP of compute (estimation method: operation counting). The model has 1,140,000,000 parameters. Training ran on 2,048 Google TPU v5e.

Access: Open weights (unrestricted). Its weights are openly available. Epoch AI rates the confidence of this record as confident.

Full record

Organization: Google DeepMind
Country of organization: United States
Domain: Vision
Task: Image classification, Image embedding
Training compute: 8.2×10²² FLOP
Compute estimation method: Operation counting
Parameters: 1,140,000,000
Training hardware: Google TPU v5e
Chips used: 2,048
Training power draw: 904.7 kW
Model accessibility: Open weights (unrestricted)
Open weights: Yes
Epoch confidence: Confident