LongBench v2 Benchmark Leaderboard | Phaseo

LongBench v2

AI benchmark results and model performance

Type: percentageGeneral

Recorded Results

17

Average Score

54.46%

Score Range

26.10% - 63.20%

Leading Model (lowest score)

26.10% - Qwen 3.5 0.8B

Scores Over Time

Individual benchmark scores plotted by date.

Models Using This Benchmark

Organisation	Model	Reported	Top Score	Info	Self Reported	Source
Qwen	Qwen 3.5 0.8B	02 Mar 2026	26.10%	-	Yes	Source
Qwen	Qwen 3.5 2B	02 Mar 2026	38.70%	-	Yes	Source
DeepSeek	DeepSeek OCR	20 Oct 2025	48.70%	inferred family alias from deepseek-v3 (score=0.3000; benches=20)	Yes	Source
DeepSeek	DeepSeek V2 (2024-06-28)	28 Jun 2024	48.70%	inferred family alias from deepseek-v3 (score=0.4159; benches=20)	Yes	Source
DeepSeek	DeepSeek V4	-	48.70%	inferred high-confidence family alias from deepseek-v3 (score=0.5818; benches=20)	Yes	Source
Qwen	Qwen 3.5 4B	02 Mar 2026	50%	-	Yes	Source
Qwen	Qwen 3.5 9B	02 Mar 2026	55.20%	-	Yes	Source
Qwen	Qwen 3.5 35B A3B	24 Feb 2026	59%	-	Yes	Source
Qwen	Qwen 3.5 122B A10B	24 Feb 2026	60.20%	-	Yes	Source
Qwen	Qwen 3.5 27B	24 Feb 2026	60.60%	-	Yes	Source
Qwen	Qwen 3.5 Flash	23 Feb 2026	60.60%	inferred family alias from qwen3.5-27b (score=0.4147; benches=81)	Yes	Source
Xiaomi	MiMo V2 Flash	16 Dec 2025	60.60%	-	Yes	Source
MiniMax	Minimax M1 40K	16 Jun 2025	61%	-	Yes	-
Moonshot	Kimi K2.5	27 Jan 2026	61%	-	Yes	Source
MiniMax	MiniMax M1 80K	16 Jun 2025	61.50%	-	Yes	-
Qwen	Qwen 3.6 Plus	01 Apr 2026	62%	-	Yes	Source
Qwen	Qwen 3.5 397B A17B	16 Feb 2026	63.20%	-	Yes	Source