Flores-200 leaderboard

As of November 14, 2026, among the results listed here, Tencent Hy-MT2-30B-A3B has the highest reported Flores-200 score, 88.1%, published by Tencent, ahead of Tencent Hy-MT2-7B at 87.1%. 4 models have a Flores-200 result in our records.

Scores as the model makers published them, not re-run by us; each row links to its announcement.

Best reported score

88.1%

Tencent Hy-MT2-30B-A3B

Reported by Tencent, August 19, 2026

Runner-up

87.1%

Tencent Hy-MT2-7B

Reported by Tencent, August 19, 2026

4 models ranked · 2 makers

Reported Flores-200 scores

Flores-200 scores reported by model makers
Rank	Model	Score	Setting	Reported by	Date
1	Hy-MT2-30B-A3B	88.1%	Not published	Tencent	August 19, 2026 (model release) · read from the maker's chart
2	Hy-MT2-7B	87.1%	Not published	Tencent	August 19, 2026 (model release) · read from the maker's chart
3	Hy-MT2-1.8B	79.2%	Not published	Tencent	August 19, 2026 (model release) · read from the maker's chart
4	H SmolLM3-3B-Base	40.2%	Base Pre-Trained Model; 5-shot; Arabic	HuggingFaceTB	Publication date not reported

Other reported runs

Flores-200 scores reported by model makers, other reported runs
Model	Score	Setting	Reported by	Date
H SmolLM3-3B-Base	29.2%	Base Pre-Trained Model; 5-shot; Chinese	HuggingFaceTB	Publication date not reported
H SmolLM3-3B-Base	62.8%	Base Pre-Trained Model; 5-shot; French	HuggingFaceTB	Publication date not reported
H SmolLM3-3B-Base	56.6%	Base Pre-Trained Model; 5-shot; German	HuggingFaceTB	Publication date not reported
H SmolLM3-3B-Base	52.6%	Base Pre-Trained Model; 5-shot; Italian	HuggingFaceTB	Publication date not reported
H SmolLM3-3B-Base	60.9%	Base Pre-Trained Model; 5-shot; Portuguese	HuggingFaceTB	Publication date not reported
H SmolLM3-3B-Base	47.1%	Base Pre-Trained Model; 5-shot; Russian	HuggingFaceTB	Publication date not reported
H SmolLM3-3B-Base	48.3%	Base Pre-Trained Model; 5-shot; Spanish	HuggingFaceTB	Publication date not reported

How much to trust these numbers

Source: Reported by the model makers themselves. Not re-run by LLMBenchmark or anyone independent.
Settings: Settings differ between makers. Check the Setting column before comparing two scores.
Spread: 88.1% to 40.2% across 4 models.
Freshness: Published August 19, 2026.
Charts: Some values were read from charts in the announcements. Two independent readings had to match exactly.

What Flores-200 measures

FLORES-200 asks models to translate text between languages, with scores indicating translation quality on a human-translated multilingual benchmark. Source: flores/flores200/README.md at main