Skip to main content
BenchmarkLLMsv2022

FLORES-200

by NLLB Team / Meta AI · open-source · Last verified 2026-03-17

FLORES-200 is a many-to-many multilingual translation benchmark covering 200 languages, including many low-resource ones. It evaluates machine translation systems across 40,000 language direction pairs, making it the most comprehensive translation benchmark for assessing cross-lingual generalization.

https://github.com/facebookresearch/flores
B+
B+Good
Adoption: AQuality: A+Freshness: B+Citations: AEngagement: F

Specifications

License
CC BY-SA 4.0
Pricing
open-source
Capabilities
evaluation, machine-translation, multilingual-evaluation
Integrations
Use Cases
model-evaluation, translation-ai, multilingual-nlp
API Available
No
Evaluated Models
nllb-200-3.3b, m2m-100, seamless-m4t, gpt-4o
Metrics
spbleu, chrf
Methodology
1,012 sentences drawn from Wikipedia, Wikinews, and Wikijourneys in a domain-balanced split. Translated by professional translators across 200 languages. spBLEU and chrF++ are the primary automatic metrics; averaged over all language directions.
Last Run
2026-01-22
Tags
translation, multilingual, low-resource, flores, spbleu
Added
2026-03-17
Completeness
100%

Index Score

72.2
Adoption
82
Quality
91
Freshness
78
Citations
85
Engagement
0

Explore the full AI ecosystem on Agents as a Service