Grok-1

Training compute

2.9×10²⁴ FLOP

Parameters

314B

Published

Nov 4, 2023

Grok-1 is an AI model developed by xAI (United States), first published in November 2023. It works in the language domain, on tasks such as language modeling and chat.

Training it took an estimated 2.9×10²⁴ FLOP of compute (estimation method: benchmarks). The model has 314,000,000,000 parameters. It was trained on roughly 6.2T datapoints.

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

Full record

Organization: xAI
Country of organization: United States
Domain: Language
Task: Language modeling, Chat
Training compute: 2.9×10²⁴ FLOP
Compute estimation method: Benchmarks
Parameters: 314,000,000,000
Dataset size: 6.2T
Model accessibility: Open weights (unrestricted)
Open weights: Yes
Epoch confidence: Likely