Llama 4 Maverick Instruct FP8 Benchmark Results

Llama 4 Maverick Instruct FP8 — benchmark results

Llama 4 Maverick Instruct evaluated in FP8 quantization. Provider: Meta. Released 2025-04-05. Access: Open.

Unified ELO 1520 ± 24, rank #724 of 1776 rated models, from 98 benchmark results.

Benchmark	Score	Metric	Percentile
HELM SeaHELM - Flores (en-ta)	51.98	ChrF++	100
HELM SeaHELM - Flores (th-en)	59.27	ChrF++	100
HELM SeaHELM - XQuAD (Thai)	77.14	SQuAD macro-averaged F1 score	100
HELM SeaHELM - IndicQA	48.42	SQuAD macro-averaged F1 score	95
HELM Capabilities - IFEval	90.76	IFEval Strict Acc	92
HELM SeaHELM - Flores (ta-en)	57.09	ChrF++	90
HELM SeaHELM - MLHSD	68.92	Macro F1 score	90
HELM SeaHELM - TyDiQA	64.9	SQuAD macro-averaged F1 score	90
CPTU Bench	3.93	Average Score (1-5)	84.2
Kluster Hallucination Detection - RAG Method 2 Resistance	96.66	Resistance (100 - Hallucination Rate %)	81.8
HELM Long Context - InfiniteBench En.MC	89	EM	80
HELM Long Context - InfiniteBench En.Sum	16.06	ROUGE-L	80