Harvey LAB-AA

Accessible data for Harvey LAB-AA leaderboard
Model	Rank	Model ID	Provider	Score	Unit
Gemini 3.7 Flash	#1	gemini-3-7-flash	Google	90.7	percent
Claude Sonnet 5	#2	claude-sonnet-5	Anthropic	90.1	percent
GPT-5.6 Terra	#3	gpt-5-6-terra	OpenAI	85.2	percent
Gemini 3.6 Flash	#4	gemini-3-6-flash	Google	85.1	percent
Kimi K3	#5	kimi-k3	Moonshot AI	26.7	percent
Grok 4.6	#6	grok-4-6	xAI	15.8	percent
Claude Fable 5	#7	claude-fable-5	Anthropic	14.1667	percent
Grok 4.5	#8	grok-4-5	xAI	13.3333	percent
Claude Opus 5	#9	claude-opus-5	Anthropic	11.7	percent
Muse Spark 1.1	#10	muse-spark-1-1	Meta	8.3333	percent
Claude Opus 4.8	#11	claude-opus-4-8	Anthropic	7.5	percent
GLM-5.2	#12	glm-5-2	Z.AI	7.5	percent
MiniMax M3	#13	minimax-m3	MiniMax	6.6667	percent
GPT-5.6 Luna	#14	gpt-5-6-luna	OpenAI	5	percent
Claude Sonnet 4.6	#15	claude-sonnet-4-6	Anthropic	4.1667	percent
GPT-5.5	#16	gpt-5-5	OpenAI	4.1667	percent
DeepSeek V4 Pro	#17	deepseek-v4-pro	DeepSeek	3.3333	percent
Nemotron 3 Ultra	#18	nemotron-3-ultra	NVIDIA	3.3333	percent
GPT-5.6 Sol	#19	gpt-5-6-sol	OpenAI	2.5	percent
DeepSeek V4 Flash	#20	deepseek-v4-flash	DeepSeek	1.6667	percent
Gemini 3.5 Flash	#21	gemini-3-5-flash	Google	1.6667	percent
Qwen3.7-Plus	#22	qwen-3-7-plus	Alibaba Cloud	1.6667	percent
Kimi K2.7 Code	#23	kimi-k2-7-code	Moonshot AI	0.8333	percent
Mistral Medium 3.5	#24	mistral-medium-3-5	Mistral AI	0.8333	percent
Claude Haiku 4.5	#25	claude-haiku-4-5	Anthropic	0	percent
Claude Opus 4.1	#26	claude-opus-4-1	Anthropic	0	percent
Gemini 3.1 Flash-Lite	#27	gemini-3-1-flash-lite	Google	0	percent
Gemini 3.1 Pro Preview	#28	gemini-3-1-pro-preview	Google	0	percent
Gemma 4 31B	#29	gemma-4-31b	Google	0	percent
GPT-5.4 mini	#30	gpt-5-4-mini	OpenAI	0	percent
GPT-5.4 nano	#31	gpt-5-4-nano	OpenAI	0	percent
Grok 4.3	#32	grok-4-3	xAI	0	percent
Kimi K2.6	#33	kimi-k2-6	Moonshot AI	0	percent
MiMo-V2.5-Pro	#34	mimo-v2-5-pro	Xiaomi	0	percent
Qwen3.5 397B A17B	#35	qwen3-5-397b	Alibaba Cloud	0	percent
Qwen3.7-Max	#36	qwen-3-7-max	Alibaba Cloud	0	percent

Harvey LAB-AA leaderboard ranked by score
Rank	Model	Provider	License	Evidence use	Score
#1	Gemini 3.7 Flash gemini-3.7-flash	Google	closed	Score input	90.7%
#2	Claude Sonnet 5 claude-sonnet-5	Anthropic	closed	Public reference	90.1%
#3	GPT-5.6 Terra gpt-5.6-terra	OpenAI	closed	Public reference	85.2%
#4	Gemini 3.6 Flash gemini-3.6-flash	Google	closed	Public reference	85.1%
#5	Kimi K3 kimi-k3	Moonshot AI	open	Score input	26.7%
#6	Grok 4.6 grok-4.6	xAI	closed	Score input	15.8%
#7	Claude Fable 5 claude-fable-5	Anthropic	closed	Score input	14.2%
#8	Grok 4.5 grok-4.5	xAI	closed	Score input	13.3%
#9	Claude Opus 5 claude-opus-5	Anthropic	closed	Score input	11.7%
#10	Muse Spark 1.1 muse-spark-1-1	Meta	closed	Score input	8.3%
#11	Claude Opus 4.8 claude-opus-4-8	Anthropic	closed	Score input	7.5%
#12	GLM-5.2 glm-5-2	Z.AI	open	Score input	7.5%
#13	MiniMax M3 MiniMax-M3	MiniMax	open	Score input	6.7%
#14	GPT-5.6 Luna gpt-5.6-luna	OpenAI	closed	Score input	5%
#15	Claude Sonnet 4.6 claude-sonnet-4-6	Anthropic	closed	Score input	4.2%
#16	GPT-5.5 gpt-5.5	OpenAI	closed	Score input	4.2%
#17	DeepSeek V4 Pro deepseek-v4-pro	DeepSeek	open	Score input	3.3%
#18	Nemotron 3 Ultra nemotron-3-ultra-550b	NVIDIA	open	Score input	3.3%
#19	GPT-5.6 Sol gpt-5.6-sol	OpenAI	closed	Relative comparison	2.5%
#20	DeepSeek V4 Flash deepseek-v4-flash	DeepSeek	open	Score input	1.7%
#21	Gemini 3.5 Flash gemini-3.5-flash	Google	closed	Score input	1.7%
#22	Qwen3.7-Plus qwen3.7-plus	Alibaba Cloud	closed	Score input	1.7%
#23	Kimi K2.7 Code kimi-k2.7-code	Moonshot AI	open	Score input	0.8%
#24	Mistral Medium 3.5 mistral-medium-3-5	Mistral AI	open	Score input	0.8%
#25	Claude Haiku 4.5 claude-haiku-4-5-20251001	Anthropic	closed	Score input	0%
#26	Claude Opus 4.1 claude-opus-4-1	Anthropic	closed	Score input	0%
#27	Gemini 3.1 Flash-Lite gemini-3.1-flash-lite	Google	closed	Score input	0%
#28	Gemini 3.1 Pro Preview gemini-3.1-pro-preview	Google	closed	Score input	0%
#29	Gemma 4 31B gemma-4-31b	Google	open	Score input	0%
#30	GPT-5.4 mini gpt-5.4-mini	OpenAI	closed	Score input	0%
#31	GPT-5.4 nano gpt-5.4-nano	OpenAI	closed	Score input	0%
#32	Grok 4.3 grok-4.3	xAI	closed	Score input	0%
#33	Kimi K2.6 kimi-k2.6	Moonshot AI	open	Score input	0%
#34	MiMo-V2.5-Pro mimo-v2-5-pro	Xiaomi	open	Score input	0%
#35	Qwen3.5 397B A17B Qwen/Qwen3.5-397B-A17B	Alibaba Cloud	open	Score input	0%
#36	Qwen3.7-Max qwen3.7-max	Alibaba Cloud	closed	Score input	0%

Rank

Model

Provider

License

Evidence use

Score

Gemini 3.7 Flash

gemini-3.7-flash

Google

closed

Score input

90.7%

Claude Sonnet 5

claude-sonnet-5

Anthropic

closed

Public reference

90.1%

GPT-5.6 Terra

gpt-5.6-terra

OpenAI

closed

Public reference

85.2%

Gemini 3.6 Flash

gemini-3.6-flash

Google

closed

Public reference

85.1%

Kimi K3

kimi-k3

Moonshot AI

open

Score input

26.7%

Grok 4.6

grok-4.6

xAI

closed

Score input

15.8%

Claude Fable 5

claude-fable-5

Anthropic

closed

Score input

14.2%

Grok 4.5

grok-4.5

xAI

closed

Score input

13.3%

Claude Opus 5

claude-opus-5

Anthropic

closed

Score input

11.7%

#10

Muse Spark 1.1

muse-spark-1-1

Every distinct dated source-published row for Harvey LAB-AA, including exact model configuration labels
Exact published model / configuration	Canonical product	Score	Version / system	Evidence	Source / observed
Gemini 3.7 Flash gemini-3-7-flash-medium	Gemini 3.7 Flash	90.7%	August 2026Artificial Analysis Harvey LAB-AA public leaderboard	Score input provider-reported	Google DeepMind permanent refresh source ↗Observed 2026-08-13 · checked 2026-08-13
Claude Sonnet 5 claude-sonnet-5-max	Claude Sonnet 5	90.1%	August 2026Artificial Analysis Harvey LAB-AA public leaderboard	Public reference · not scored provider-reported	Google DeepMind permanent refresh source ↗Observed 2026-08-13 · checked 2026-08-13
GPT-5.6 Terra gpt-5-6-terra-max	GPT-5.6 Terra	85.2%	August 2026Artificial Analysis Harvey LAB-AA public leaderboard	Public reference · not scored provider-reported	Google DeepMind permanent refresh source ↗Observed 2026-08-13 · checked 2026-08-13
Gemini 3.6 Flash gemini-3-6-flash-high	Gemini 3.6 Flash	85.1%	August 2026Artificial Analysis Harvey LAB-AA public leaderboard	Public reference · not scored provider-reported	Google DeepMind permanent refresh source ↗Observed 2026-08-13 · checked 2026-08-13
Grok 4.6 High grok-4-6-high	Grok 4.6	15.8%	CurrentxAI Grok 4.6 official release comparison table	Score input provider-reported	Grok 4.6 ↗Observed 2026-08-12 · checked 2026-08-12
Grok 4.5 High grok-4-5-aa-2-high	Grok 4.5	12.9%	CurrentxAI Grok 4.6 official release comparison table	Relative comparison · not standard direct provider-reported	Grok 4.6 ↗Observed 2026-08-12 · checked 2026-08-12
Fable 5 Max claude-fable-5-max	Claude Fable 5	11.3%	CurrentxAI Grok 4.6 official release comparison table	Relative comparison · not standard direct provider-reported	Grok 4.6 ↗Observed 2026-08-12 · checked 2026-08-12
GPT-5.6 Sol Max gpt-5-6-sol-max	GPT-5.6 Sol	2.5%	CurrentxAI Grok 4.6 official release comparison table	Relative comparison · not standard direct provider-reported	Grok 4.6 ↗Observed 2026-08-12 · checked 2026-08-12
Legal Agent Benchmark held-out split as published in the Anthropic Claude Opus 5 launch table. source label without registered configuration ID	Claude Opus 5	11.7%	CurrentAnthropic provider evaluation	Score input provider-reported	Introducing Claude Opus 5 ↗Observed 2026-07-24 · checked 2026-07-24
Kimi K3; Artificial Analysis Harvey LAB-AA evaluation. source label without registered configuration ID	Kimi K3	26.7%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Kimi K3 Intelligence, Performance & Price Analysis ↗Observed 2026-07-20 · checked 2026-07-21
Muse Spark 1.1 (xhigh); Artificial Analysis independent evaluation configuration as published. muse-spark-1-1-xhigh	Muse Spark 1.1	8.3%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Muse Spark 1.1 (xhigh) analysis ↗Observed 2026-07-16 · checked 2026-07-16
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) claude-fable-5-max	Claude Fable 5	14.2%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Grok 4.5 (high) source label without registered configuration ID	Grok 4.5	13.3%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Muse Spark 1.1 (xhigh) muse-spark-1-1-xhigh	Muse Spark 1.1	8.3%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Claude Opus 4.8 (Adaptive Reasoning, Max Effort) claude-opus-4-8-max	Claude Opus 4.8	7.5%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
GLM-5.2 (max) glm-5-2-max	GLM-5.2	7.5%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
MiniMax-M3 source label without registered configuration ID	MiniMax M3	6.7%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
GPT-5.6 Luna (max) gpt-5-6-luna-max	GPT-5.6 Luna	5%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Claude Sonnet 5 (Adaptive Reasoning, Max Effort) claude-sonnet-5-max	Claude Sonnet 5	5%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) claude-sonnet-4-6-max	Claude Sonnet 4.6	4.2%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
GPT-5.5 (xhigh) gpt-5-5-xhigh	GPT-5.5	4.2%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
DeepSeek V4 Pro (Reasoning, Max Effort) deepseek-v4-pro-max	DeepSeek V4 Pro	3.3%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Nemotron 3 Ultra 550B A55B (Reasoning) source label without registered configuration ID	Nemotron 3 Ultra	3.3%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
GPT-5.6 Terra (max) gpt-5-6-terra-max	GPT-5.6 Terra	2.5%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
DeepSeek V4 Flash (Reasoning, Max Effort) deepseek-v4-flash-max	DeepSeek V4 Flash	1.7%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Qwen3.7 Plus source label without registered configuration ID	Qwen3.7-Plus	1.7%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Gemini 3.5 Flash (high) gemini-3-5-flash-high	Gemini 3.5 Flash	1.7%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Mistral Medium 3.5 source label without registered configuration ID	Mistral Medium 3.5	0.8%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Kimi K2.7 Code source label without registered configuration ID	Kimi K2.7 Code	0.8%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Gemini 3.1 Flash-Lite source label without registered configuration ID	Gemini 3.1 Flash-Lite	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Grok 4.3 (high) source label without registered configuration ID	Grok 4.3	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Qwen3.7 Max source label without registered configuration ID	Qwen3.7-Max	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
GPT-5.4 nano (xhigh) gpt-5-4-nano-xhigh	GPT-5.4 nano	0%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Claude 4.5 Haiku (Reasoning) source label without registered configuration ID	Claude Haiku 4.5	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Gemini 3.1 Pro Preview source label without registered configuration ID	Gemini 3.1 Pro Preview	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Gemma 4 31B (Reasoning) source label without registered configuration ID	Gemma 4 31B	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
GPT-5.4 mini (xhigh) gpt-5-4-mini-xhigh	GPT-5.4 mini	0%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Claude 4.1 Opus (Reasoning) source label without registered configuration ID	Claude Opus 4.1	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Kimi K2.6 source label without registered configuration ID	Kimi K2.6	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Qwen3.5 397B A17B (Reasoning) source label without registered configuration ID	Qwen3.5 397B A17B	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
MiMo-V2.5-Pro source label without registered configuration ID	MiMo-V2.5-Pro	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15

What does Harvey LAB-AA measure?

Legal agent benchmark on private Harvey legal work tasks graded criterion-by-criterion in an agent sandbox.

Which model scores highest on Harvey LAB-AA?

Gemini 3.7 Flash by Google currently leads with 90.7%.

How many models are evaluated on Harvey LAB-AA?

36 canonical model products have a qualifying score, backed by 41 distinct dated source rows.

Does this affect overall Lumina rank?

Yes. This family is ranking-weighted at 0.9% of its capability category. Gemini 3.7 Flash is overall #10.

Accessible data for Harvey LAB-AA leaderboard
Model	Rank	Model ID	Provider	Score	Unit
Gemini 3.7 Flash	#1	gemini-3-7-flash	Google	90.7	percent
Claude Sonnet 5	#2	claude-sonnet-5	Anthropic	90.1	percent
GPT-5.6 Terra	#3	gpt-5-6-terra	OpenAI	85.2	percent
Gemini 3.6 Flash	#4	gemini-3-6-flash	Google	85.1	percent
Kimi K3	#5	kimi-k3	Moonshot AI	26.7	percent
Grok 4.6	#6	grok-4-6	xAI	15.8	percent
Claude Fable 5	#7	claude-fable-5	Anthropic	14.1667	percent
Grok 4.5	#8	grok-4-5	xAI	13.3333	percent
Claude Opus 5	#9	claude-opus-5	Anthropic	11.7	percent
Muse Spark 1.1	#10	muse-spark-1-1	Meta	8.3333	percent
Claude Opus 4.8	#11	claude-opus-4-8	Anthropic	7.5	percent
GLM-5.2	#12	glm-5-2	Z.AI	7.5	percent
MiniMax M3	#13	minimax-m3	MiniMax	6.6667	percent
GPT-5.6 Luna	#14	gpt-5-6-luna	OpenAI	5	percent
Claude Sonnet 4.6	#15	claude-sonnet-4-6	Anthropic	4.1667	percent
GPT-5.5	#16	gpt-5-5	OpenAI	4.1667	percent
DeepSeek V4 Pro	#17	deepseek-v4-pro	DeepSeek	3.3333	percent
Nemotron 3 Ultra	#18	nemotron-3-ultra	NVIDIA	3.3333	percent
GPT-5.6 Sol	#19	gpt-5-6-sol	OpenAI	2.5	percent
DeepSeek V4 Flash	#20	deepseek-v4-flash	DeepSeek	1.6667	percent
Gemini 3.5 Flash	#21	gemini-3-5-flash	Google	1.6667	percent
Qwen3.7-Plus	#22	qwen-3-7-plus	Alibaba Cloud	1.6667	percent
Kimi K2.7 Code	#23	kimi-k2-7-code	Moonshot AI	0.8333	percent
Mistral Medium 3.5	#24	mistral-medium-3-5	Mistral AI	0.8333	percent
Claude Haiku 4.5	#25	claude-haiku-4-5	Anthropic	0	percent
Claude Opus 4.1	#26	claude-opus-4-1	Anthropic	0	percent
Gemini 3.1 Flash-Lite	#27	gemini-3-1-flash-lite	Google	0	percent
Gemini 3.1 Pro Preview	#28	gemini-3-1-pro-preview	Google	0	percent
Gemma 4 31B	#29	gemma-4-31b	Google	0	percent
GPT-5.4 mini	#30	gpt-5-4-mini	OpenAI	0	percent
GPT-5.4 nano	#31	gpt-5-4-nano	OpenAI	0	percent
Grok 4.3	#32	grok-4-3	xAI	0	percent
Kimi K2.6	#33	kimi-k2-6	Moonshot AI	0	percent
MiMo-V2.5-Pro	#34	mimo-v2-5-pro	Xiaomi	0	percent
Qwen3.5 397B A17B	#35	qwen3-5-397b	Alibaba Cloud	0	percent
Qwen3.7-Max	#36	qwen-3-7-max	Alibaba Cloud	0	percent

Harvey LAB-AA leaderboard ranked by score
Rank	Model	Provider	License	Evidence use	Score
#1	Gemini 3.7 Flash gemini-3.7-flash	Google	closed	Score input	90.7%
#2	Claude Sonnet 5 claude-sonnet-5	Anthropic	closed	Public reference	90.1%
#3	GPT-5.6 Terra gpt-5.6-terra	OpenAI	closed	Public reference	85.2%
#4	Gemini 3.6 Flash gemini-3.6-flash	Google	closed	Public reference	85.1%
#5	Kimi K3 kimi-k3	Moonshot AI	open	Score input	26.7%
#6	Grok 4.6 grok-4.6	xAI	closed	Score input	15.8%
#7	Claude Fable 5 claude-fable-5	Anthropic	closed	Score input	14.2%
#8	Grok 4.5 grok-4.5	xAI	closed	Score input	13.3%
#9	Claude Opus 5 claude-opus-5	Anthropic	closed	Score input	11.7%
#10	Muse Spark 1.1 muse-spark-1-1	Meta	closed	Score input	8.3%
#11	Claude Opus 4.8 claude-opus-4-8	Anthropic	closed	Score input	7.5%
#12	GLM-5.2 glm-5-2	Z.AI	open	Score input	7.5%
#13	MiniMax M3 MiniMax-M3	MiniMax	open	Score input	6.7%
#14	GPT-5.6 Luna gpt-5.6-luna	OpenAI	closed	Score input	5%
#15	Claude Sonnet 4.6 claude-sonnet-4-6	Anthropic	closed	Score input	4.2%
#16	GPT-5.5 gpt-5.5	OpenAI	closed	Score input	4.2%
#17	DeepSeek V4 Pro deepseek-v4-pro	DeepSeek	open	Score input	3.3%
#18	Nemotron 3 Ultra nemotron-3-ultra-550b	NVIDIA	open	Score input	3.3%
#19	GPT-5.6 Sol gpt-5.6-sol	OpenAI	closed	Relative comparison	2.5%
#20	DeepSeek V4 Flash deepseek-v4-flash	DeepSeek	open	Score input	1.7%
#21	Gemini 3.5 Flash gemini-3.5-flash	Google	closed	Score input	1.7%
#22	Qwen3.7-Plus qwen3.7-plus	Alibaba Cloud	closed	Score input	1.7%
#23	Kimi K2.7 Code kimi-k2.7-code	Moonshot AI	open	Score input	0.8%
#24	Mistral Medium 3.5 mistral-medium-3-5	Mistral AI	open	Score input	0.8%
#25	Claude Haiku 4.5 claude-haiku-4-5-20251001	Anthropic	closed	Score input	0%
#26	Claude Opus 4.1 claude-opus-4-1	Anthropic	closed	Score input	0%
#27	Gemini 3.1 Flash-Lite gemini-3.1-flash-lite	Google	closed	Score input	0%
#28	Gemini 3.1 Pro Preview gemini-3.1-pro-preview	Google	closed	Score input	0%
#29	Gemma 4 31B gemma-4-31b	Google	open	Score input	0%
#30	GPT-5.4 mini gpt-5.4-mini	OpenAI	closed	Score input	0%
#31	GPT-5.4 nano gpt-5.4-nano	OpenAI	closed	Score input	0%
#32	Grok 4.3 grok-4.3	xAI	closed	Score input	0%
#33	Kimi K2.6 kimi-k2.6	Moonshot AI	open	Score input	0%
#34	MiMo-V2.5-Pro mimo-v2-5-pro	Xiaomi	open	Score input	0%
#35	Qwen3.5 397B A17B Qwen/Qwen3.5-397B-A17B	Alibaba Cloud	open	Score input	0%
#36	Qwen3.7-Max qwen3.7-max	Alibaba Cloud	closed	Score input	0%

Rank

Model

Provider

License

Evidence use

Score

Gemini 3.7 Flash

gemini-3.7-flash

Google

closed

Score input

90.7%

Claude Sonnet 5

claude-sonnet-5

Anthropic

closed

Public reference

90.1%

GPT-5.6 Terra

gpt-5.6-terra

OpenAI

closed

Public reference

85.2%

Gemini 3.6 Flash

gemini-3.6-flash

Google

closed

Public reference

85.1%

Kimi K3

kimi-k3

Moonshot AI

open

Score input

26.7%

Grok 4.6

grok-4.6

xAI

closed

Score input

15.8%

Claude Fable 5

claude-fable-5

Anthropic

closed

Score input

14.2%

Grok 4.5

grok-4.5

xAI

closed

Score input

13.3%

Claude Opus 5

claude-opus-5

Anthropic

closed

Score input

11.7%

#10

Muse Spark 1.1

muse-spark-1-1

Every distinct dated source-published row for Harvey LAB-AA, including exact model configuration labels
Exact published model / configuration	Canonical product	Score	Version / system	Evidence	Source / observed
Gemini 3.7 Flash gemini-3-7-flash-medium	Gemini 3.7 Flash	90.7%	August 2026Artificial Analysis Harvey LAB-AA public leaderboard	Score input provider-reported	Google DeepMind permanent refresh source ↗Observed 2026-08-13 · checked 2026-08-13
Claude Sonnet 5 claude-sonnet-5-max	Claude Sonnet 5	90.1%	August 2026Artificial Analysis Harvey LAB-AA public leaderboard	Public reference · not scored provider-reported	Google DeepMind permanent refresh source ↗Observed 2026-08-13 · checked 2026-08-13
GPT-5.6 Terra gpt-5-6-terra-max	GPT-5.6 Terra	85.2%	August 2026Artificial Analysis Harvey LAB-AA public leaderboard	Public reference · not scored provider-reported	Google DeepMind permanent refresh source ↗Observed 2026-08-13 · checked 2026-08-13
Gemini 3.6 Flash gemini-3-6-flash-high	Gemini 3.6 Flash	85.1%	August 2026Artificial Analysis Harvey LAB-AA public leaderboard	Public reference · not scored provider-reported	Google DeepMind permanent refresh source ↗Observed 2026-08-13 · checked 2026-08-13
Grok 4.6 High grok-4-6-high	Grok 4.6	15.8%	CurrentxAI Grok 4.6 official release comparison table	Score input provider-reported	Grok 4.6 ↗Observed 2026-08-12 · checked 2026-08-12
Grok 4.5 High grok-4-5-aa-2-high	Grok 4.5	12.9%	CurrentxAI Grok 4.6 official release comparison table	Relative comparison · not standard direct provider-reported	Grok 4.6 ↗Observed 2026-08-12 · checked 2026-08-12
Fable 5 Max claude-fable-5-max	Claude Fable 5	11.3%	CurrentxAI Grok 4.6 official release comparison table	Relative comparison · not standard direct provider-reported	Grok 4.6 ↗Observed 2026-08-12 · checked 2026-08-12
GPT-5.6 Sol Max gpt-5-6-sol-max	GPT-5.6 Sol	2.5%	CurrentxAI Grok 4.6 official release comparison table	Relative comparison · not standard direct provider-reported	Grok 4.6 ↗Observed 2026-08-12 · checked 2026-08-12
Legal Agent Benchmark held-out split as published in the Anthropic Claude Opus 5 launch table. source label without registered configuration ID	Claude Opus 5	11.7%	CurrentAnthropic provider evaluation	Score input provider-reported	Introducing Claude Opus 5 ↗Observed 2026-07-24 · checked 2026-07-24
Kimi K3; Artificial Analysis Harvey LAB-AA evaluation. source label without registered configuration ID	Kimi K3	26.7%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Kimi K3 Intelligence, Performance & Price Analysis ↗Observed 2026-07-20 · checked 2026-07-21
Muse Spark 1.1 (xhigh); Artificial Analysis independent evaluation configuration as published. muse-spark-1-1-xhigh	Muse Spark 1.1	8.3%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Muse Spark 1.1 (xhigh) analysis ↗Observed 2026-07-16 · checked 2026-07-16
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) claude-fable-5-max	Claude Fable 5	14.2%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Grok 4.5 (high) source label without registered configuration ID	Grok 4.5	13.3%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Muse Spark 1.1 (xhigh) muse-spark-1-1-xhigh	Muse Spark 1.1	8.3%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Claude Opus 4.8 (Adaptive Reasoning, Max Effort) claude-opus-4-8-max	Claude Opus 4.8	7.5%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
GLM-5.2 (max) glm-5-2-max	GLM-5.2	7.5%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
MiniMax-M3 source label without registered configuration ID	MiniMax M3	6.7%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
GPT-5.6 Luna (max) gpt-5-6-luna-max	GPT-5.6 Luna	5%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Claude Sonnet 5 (Adaptive Reasoning, Max Effort) claude-sonnet-5-max	Claude Sonnet 5	5%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) claude-sonnet-4-6-max	Claude Sonnet 4.6	4.2%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
GPT-5.5 (xhigh) gpt-5-5-xhigh	GPT-5.5	4.2%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
DeepSeek V4 Pro (Reasoning, Max Effort) deepseek-v4-pro-max	DeepSeek V4 Pro	3.3%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Nemotron 3 Ultra 550B A55B (Reasoning) source label without registered configuration ID	Nemotron 3 Ultra	3.3%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
GPT-5.6 Terra (max) gpt-5-6-terra-max	GPT-5.6 Terra	2.5%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
DeepSeek V4 Flash (Reasoning, Max Effort) deepseek-v4-flash-max	DeepSeek V4 Flash	1.7%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Qwen3.7 Plus source label without registered configuration ID	Qwen3.7-Plus	1.7%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Gemini 3.5 Flash (high) gemini-3-5-flash-high	Gemini 3.5 Flash	1.7%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Mistral Medium 3.5 source label without registered configuration ID	Mistral Medium 3.5	0.8%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Kimi K2.7 Code source label without registered configuration ID	Kimi K2.7 Code	0.8%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Gemini 3.1 Flash-Lite source label without registered configuration ID	Gemini 3.1 Flash-Lite	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Grok 4.3 (high) source label without registered configuration ID	Grok 4.3	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Qwen3.7 Max source label without registered configuration ID	Qwen3.7-Max	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
GPT-5.4 nano (xhigh) gpt-5-4-nano-xhigh	GPT-5.4 nano	0%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Claude 4.5 Haiku (Reasoning) source label without registered configuration ID	Claude Haiku 4.5	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Gemini 3.1 Pro Preview source label without registered configuration ID	Gemini 3.1 Pro Preview	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Gemma 4 31B (Reasoning) source label without registered configuration ID	Gemma 4 31B	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
GPT-5.4 mini (xhigh) gpt-5-4-mini-xhigh	GPT-5.4 mini	0%	CurrentArtificial Analysis independent evaluation	Public reference · not scored source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Claude 4.1 Opus (Reasoning) source label without registered configuration ID	Claude Opus 4.1	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Kimi K2.6 source label without registered configuration ID	Kimi K2.6	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
Qwen3.5 397B A17B (Reasoning) source label without registered configuration ID	Qwen3.5 397B A17B	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15
MiMo-V2.5-Pro source label without registered configuration ID	MiMo-V2.5-Pro	0%	CurrentArtificial Analysis independent evaluation	Score input source-checked	Harvey LAB-AA Leaderboard ↗Observed 2026-07-15 · checked 2026-07-15

What does Harvey LAB-AA measure?

Legal agent benchmark on private Harvey legal work tasks graded criterion-by-criterion in an agent sandbox.

Which model scores highest on Harvey LAB-AA?

Gemini 3.7 Flash by Google currently leads with 90.7%.

How many models are evaluated on Harvey LAB-AA?

36 canonical model products have a qualifying score, backed by 41 distinct dated source rows.

Does this affect overall Lumina rank?

Yes. This family is ranking-weighted at 0.9% of its capability category. Gemini 3.7 Flash is overall #10.

Navigation

Harvey LAB-AA

Leaderboard · 36 models

Gemini 3.7 Flash

Claude Sonnet 5

GPT-5.6 Terra

Harvey LAB-AA leaderboard

Crop Harvey LAB-AA leaderboard

Benchmark score table (36 models)

Crop Harvey LAB-AA benchmark table

Crop Harvey LAB-AA published source rows

Compare top models

Definition and scoring

Quick answers

What does Harvey LAB-AA measure?

Which model scores highest on Harvey LAB-AA?

How many models are evaluated on Harvey LAB-AA?

Does this affect overall Lumina rank?

Harvey LAB-AA

Leaderboard · 36 models

Gemini 3.7 Flash

Claude Sonnet 5

GPT-5.6 Terra

Harvey LAB-AA leaderboard

Crop Harvey LAB-AA leaderboard

Benchmark score table (36 models)

Crop Harvey LAB-AA benchmark table

Crop Harvey LAB-AA published source rows

Compare top models

Definition and scoring

Quick answers

What does Harvey LAB-AA measure?

Which model scores highest on Harvey LAB-AA?

How many models are evaluated on Harvey LAB-AA?

Does this affect overall Lumina rank?

Leaderboard · 36 models

Gemini 3.7 Flash

Claude Sonnet 5

GPT-5.6 Terra

Benchmark score table (36 models)

Compare top models

Definition and scoring

Quick answers

What does Harvey LAB-AA measure?

Which model scores highest on Harvey LAB-AA?

How many models are evaluated on Harvey LAB-AA?

Does this affect overall Lumina rank?

More Agents benchmarks

Leaderboard · 36 models

Gemini 3.7 Flash

Claude Sonnet 5

GPT-5.6 Terra

Benchmark score table (36 models)

Compare top models

Definition and scoring

Quick answers

What does Harvey LAB-AA measure?

Which model scores highest on Harvey LAB-AA?

How many models are evaluated on Harvey LAB-AA?

Does this affect overall Lumina rank?

More Agents benchmarks