AutoBench Run 5 - December 2025

Latest AutoBench run with models Gpt 5.2, Claude Opus 4.5, DeepSeek 3.2 Speciale and more

Latest

Date

December 16, 2025

Version

2025-12-16

Models

New Models

Run data

Overall Comparison Costs Latency P99 Latency Domain

Comparison of AutoBench scores with other popular benchmarks. AutoBench features 87.08% correlation with Artificial Analysis Intelligence Index, 77.16% with LMArena (Chatbot Arena), and 80.68% with MMLU-Plus. Models sorted by AutoBench score.

Model	AutoBench	Chatbot Ar.	AAI Index	MMLU Index
Gpt-5.2-pro	4.48 (#1)	-	72 (#2)	0.87 (#4)
Gpt-5.2	4.43 (#2)	-	-	-
gemini-3-pro-preview	4.41 (#3)	1492 (#1)	73 (#1)	0.9 (#1)
Claude-opus-4.5	4.39 (#4)	1470 (#3)	70 (#3)	0.9 (#2)
gpt-5.1	4.38 (#5)	1457 (#4)	70 (#4)	0.87 (#5)
Kimi-K2-thinking	4.32 (#6)	1429 (#7)	67 (#5)	0.85 (#9)
claude-sonnet-4.5	4.3 (#7)	1450 (#6)	63 (#9)	0.88 (#3)
gemini-2.5-pro	4.29 (#8)	1451 (#5)	60 (#12)	0.86 (#7)
gpt-5-mini	4.29 (#9)	1392 (#18)	64 (#8)	0.84 (#12)
grok-4.1-fast-thinking	4.21 (#10)	-	64 (#7)	0.85 (#10)
Qwen3-235B-A22B-Thinking-2507	4.2 (#12)	1397 (#16)	57 (#14)	0.84 (#13)
grok-4	4.2 (#11)	1478 (#2)	65 (#6)	0.87 (#6)
gpt-oss-120b	4.18 (#13)	1352 (#23)	61 (#11)	0.81 (#22)
claude-haiku-4.5	4.17 (#15)	1402 (#15)	55 (#16)	0.76 (#28)
gemini-2.5-flash	4.17 (#14)	1408 (#14)	51 (#21)	0.84 (#14)
Deepseek-v3.2-speciale	4.14 (#16)	1418 (#9)	59 (#13)	0.86 (#8)
GLM-4.6	4.13 (#17)	1425 (#8)	56 (#15)	0.83 (#16)
deepSeek-R1-0528	4.12 (#18)	1395 (#17)	52 (#18)	0.85 (#11)
Deepseek-v3.2	4.11 (#20)	1414 (#12)	52 (#19)	0.84 (#15)
Kimi-K2-0905	4.11 (#19)	1416 (#10)	50 (#23)	0.82 (#18)
Nova-2-lite-v1	4.06 (#22)	1334 (#27)	47 (#25)	0.81 (#23)
gpt-5-nano	4.06 (#21)	1339 (#26)	51 (#22)	0.77 (#27)
Qwen3-next-80b-a3b-thinking	4.03 (#23)	1367 (#22)	54 (#17)	0.82 (#19)
Minimax-m2	3.99 (#24)	1345 (#24)	61 (#10)	0.82 (#20)
Qwen3-235B-A22B-2507	3.98 (#25)	1374 (#20)	45 (#26)	0.83 (#17)
gemini-2.5-flash-lite	3.95 (#26)	1378 (#19)	40 (#28)	0.81 (#24)
Mistral-large-2512	3.94 (#27)	1415 (#11)	38 (#29)	0.81 (#25)
grok-4.1-fast	3.88 (#28)	-	38 (#30)	0.74 (#30)
GLM-4.5-Air	3.86 (#29)	1370 (#21)	49 (#24)	0.82 (#21)
Mistral-medium-3.1	3.81 (#30)	1411 (#13)	35 (#32)	0.68 (#33)
Gpt-oss-20b	3.78 (#32)	1318 (#28)	52 (#20)	0.75 (#29)
llama-3.3-nemotron-super-49b-v1.5	3.78 (#31)	1340 (#25)	45 (#27)	0.81 (#26)
Ministral-8b-2512	3.57 (#33)	-	28 (#34)	0.64 (#34)
nemotron-nano-9b-v2	3.5 (#34)	-	37 (#31)	0.74 (#31)
nova-premier-v1	3.47 (#35)	-	32 (#33)	0.73 (#32)