Live
Head to head

Claude 3.5 Sonnet vs GPT-4 (Mar 2023)

Anthropic
Claude 3.5 Sonnet
June 2024
vs
OpenAI
GPT-4 (Mar 2023)
March 2023
2.7×10²⁵Training compute (FLOP)2.1×10²⁵
130Capability index125.9
$26MTraining cost$37M
Benchmarks won (1 scored on both)
1 Claude 3.5 SonnetGPT-4 (Mar 2023) 0

Claude 3.5 Sonnet (Anthropic) and GPT-4 (Mar 2023) (OpenAI) are both frontier AI models. Claude 3.5 Sonnet was published in June 2024 and GPT-4 (Mar 2023) in March 2023.

Claude 3.5 Sonnet was trained on 2.7×10²⁵ FLOP, about 1.3x the compute of GPT-4 (Mar 2023) at 2.1×10²⁵ FLOP. Training compute is the closest available proxy for how much was invested in a model, though it says nothing on its own about how well that compute was spent.

Claude 3.5 Sonnet wins the single benchmark both were scored on. Benchmark counts are a crude scoreboard — the evaluations differ wildly in what they measure and in how saturated they are — so the per-benchmark table below matters more than the tally.

Specifications
Anthropic
Organization
OpenAI
Jun 20, 2024
Published
Mar 15, 2023
2.7×10²⁵ FLOP
Training compute1.3x
2.1×10²⁵ FLOP
--
Parameters
1.8T
--
Dataset size
5.4T
--
Training hardware
NVIDIA A100 SXM4 40 GB
--
Chips used
25,000
--
Training time
2,280 h
$26M
Training cost (2023 USD)1.4x
$37M
--
Training power draw
19.9 MW
API access
Accessibility
API access
No
Open weights
No
United States
Country
United States
Shared benchmarks
Related comparisons
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.