En vivo
Head to head

Falcon-180B vs Llama 4 Behemoth (preview)

Technology Innovation Institute
Falcon-180B
September 2023
vs
3.8×10²⁴Training compute (FLOP)5.2×10²⁵
$11MTraining cost$45M

Falcon-180B (Technology Innovation Institute) and Llama 4 Behemoth (preview) (Meta AI) are both frontier AI models. Falcon-180B was published in September 2023 and Llama 4 Behemoth (preview) in April 2025.

Llama 4 Behemoth (preview) was trained on 5.2×10²⁵ FLOP, about 13.8x the compute of Falcon-180B at 3.8×10²⁴ 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.

These two models share no benchmark on which both have been scored, so no direct performance comparison is possible here. The specification table below is a comparison of inputs, not of results.

Specifications
Technology Innovation Institute
Organization
Meta AI
Sep 6, 2023
Published
Apr 5, 2025
3.8×10²⁴ FLOP
Training compute14x
5.2×10²⁵ FLOP
180B
Parameters11x
2T
3.5T
Dataset size8.6x
30T
NVIDIA A100 SXM4 40 GB
Training hardware
NVIDIA H100 SXM5 80GB
4,096
Chips used7.8x
32,000
4,320 h
Training time
--
$11M
Training cost (2023 USD)4.2x
$45M
3.3 MW
Training power draw
43.9 MW
Open weights (restricted use)
Accessibility
Unreleased
Yes
Open weights
No
United Arab Emirates
Country
United States
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.