Agents · Benchmark profile

GDPval-AA v2

Artificial Analysis agentic evaluation of OpenAI GDPval real-world economic work tasks across occupations, scored with pairwise Elo then normalized.

Ranking weightedv2Active66 models

Data verified 15 July 2026 · Methodology 1.2.0

Benchmark score on GDPval-AA v2

Leaderboard · 66 models

Claude Fable 5 leads at 63.0%, followed by GPT-5.6 Sol (62.4%) and Kimi K3 (59.2%).

Ranking · 1.7%Agentsv2

#1Closed

Claude Fable 5

Anthropic

claude-fable-5

63.0%

Overall #2 · Context 1M tokens

#2Closed

GPT-5.6 Sol

OpenAI

gpt-5.6-sol

62.4%

Overall #4 · Context 1.05M tokens

#3Closed

Kimi K3

Moonshot AI

kimi-k3

59.2%

Overall #3 · Context 1M tokens

Visual analysis

See the field, not just the table.

Three views of the same sourced leaderboard: model placement, the score curve, and descriptive provider averages. These visuals never alter ranking calculations.

ModelsColours

AnthropicOpenAIMoonshot AIxAIZ.AIMiniMaxMetaGoogleDeepSeekXiaomiAlibaba Cloud

Leaderboard bars

20 ranked models · higher is better

View accessible chart data

Accessible data for Leaderboard bars
Model	Rank	Provider	Score
Claude Fable 5	#1	Anthropic	63.0%
GPT-5.6 Sol	#2	OpenAI	62.4%
Kimi K3	#3	Moonshot AI	59.2%
Claude Opus 4	#4	Anthropic	56.5%
Claude Sonnet 5	#5	Anthropic	55.3%
Claude Opus 4.8	#6	Anthropic	55.0%
GPT-5.6 Terra	#7	OpenAI	54.6%
GPT-5.6 Luna	#8	OpenAI	54.6%
Grok 4.5	#9	xAI	51.7%
GLM-5.2	#10	Z.AI	50.7%
Claude Opus 4.7	#11	Anthropic	50.0%
GPT-5.5	#12	OpenAI	49.7%
MiniMax M3	#13	MiniMax	44.7%
GPT-5.4	#14	OpenAI	44.7%
Claude Sonnet 4.6	#15	Anthropic	43.9%
Muse Spark 1.1	#16	Meta	43.7%
Gemini 3.5 Flash	#17	Google	42.4%
DeepSeek V4 Pro	#18	DeepSeek	40.4%
MiMo-V2-Flash	#19	Xiaomi	40.4%
Qwen3.7-Max	#20	Alibaba Cloud	38.7%

Score curve

Shows how quickly scores separate down the field

View accessible chart data

Accessible data for Score curve
Model	Rank	Provider	Score
Claude Fable 5	#1	Anthropic	63.0%
GPT-5.6 Sol	#2	OpenAI	62.4%
Kimi K3	#3	Moonshot AI	59.2%
Claude Opus 4	#4	Anthropic	56.5%
Claude Sonnet 5	#5	Anthropic	55.3%
Claude Opus 4.8	#6	Anthropic	55.0%
GPT-5.6 Terra	#7	OpenAI	54.6%
GPT-5.6 Luna	#8	OpenAI	54.6%
Grok 4.5	#9	xAI	51.7%
GLM-5.2	#10	Z.AI	50.7%
Claude Opus 4.7	#11	Anthropic	50.0%
GPT-5.5	#12	OpenAI	49.7%
MiniMax M3	#13	MiniMax	44.7%
GPT-5.4	#14	OpenAI	44.7%
Claude Sonnet 4.6	#15	Anthropic	43.9%
Muse Spark 1.1	#16	Meta	43.7%
Gemini 3.5 Flash	#17	Google	42.4%
DeepSeek V4 Pro	#18	DeepSeek	40.4%
MiMo-V2-Flash	#19	Xiaomi	40.4%
Qwen3.7-Max	#20	Alibaba Cloud	38.7%

Provider profile

Average within the visible cohort · descriptive only

View accessible chart data

Accessible data for Provider profile
Provider	Models	Average score
Moonshot AI	1	59.2%
Anthropic	6	53.9%
OpenAI	5	53.2%
xAI	1	51.7%
Z.AI	1	50.7%
MiniMax	1	44.7%
Meta	1	43.7%
Google	1	42.4%
DeepSeek	1	40.4%
Xiaomi	1	40.4%
Alibaba Cloud	1	38.7%

Benchmark score table (66 models)

One best score per model · higher is better

GDPval-AA v2 leaderboard ranked by score
Rank	Model	Provider	License	Score
#1	Claude Fable 5 claude-fable-5	Anthropic	closed	63.0%
#2	GPT-5.6 Sol gpt-5.6-sol	OpenAI	closed	62.4%
#3	Kimi K3 kimi-k3	Moonshot AI	closed	59.2%
#4	Claude Opus 4 claude-opus-4	Anthropic	closed	56.5%
#5	Claude Sonnet 5 claude-sonnet-5	Anthropic	closed	55.3%
#6	Claude Opus 4.8 claude-opus-4-8	Anthropic	closed	55.0%
#7	GPT-5.6 Terra gpt-5.6-terra	OpenAI	closed	54.6%
#8	GPT-5.6 Luna gpt-5.6-luna	OpenAI	closed	54.6%
#9	Grok 4.5 grok-4.5	xAI	closed	51.7%
#10	GLM-5.2 glm-5-2	Z.AI	open	50.7%
#11	Claude Opus 4.7 claude-opus-4-7	Anthropic	closed	50.0%
#12	GPT-5.5 gpt-5.5	OpenAI	closed	49.7%
#13	MiniMax M3 MiniMax-M3	MiniMax	—	44.7%
#14	GPT-5.4 gpt-5.4	OpenAI	closed	44.7%
#15	Claude Sonnet 4.6 claude-sonnet-4-6	Anthropic	closed	43.9%
#16	Muse Spark 1.1 muse-spark-1-1	Meta	closed	43.7%
#17	Gemini 3.5 Flash gemini-3.5-flash	Google	closed	42.4%
#18	DeepSeek V4 Pro deepseek-v4-pro	DeepSeek	open	40.4%
#19	MiMo-V2-Flash mimo-v2-flash	Xiaomi	open	40.4%
#20	Qwen3.7-Max qwen3.7-max	Alibaba Cloud	—	38.7%
#21	MiMo-V2.5-Pro mimo-v2-5-pro	Xiaomi	closed	38.3%
#22	GLM-5.1 glm-5-1	Z.AI	open	37.8%
#23	Claude Sonnet 3.7 claude-3-7-sonnet	Anthropic	closed	35.9%
#24	Hy3 hy3	Tencent	open	35.7%
#25	Kimi K2.6 kimi-k2.6	Moonshot AI	open	34.5%
#26	DeepSeek V4 Flash deepseek-v4-flash	DeepSeek	open	34.4%
#27	Kimi K2.7 Code kimi-k2.7-code	Moonshot AI	open	34.3%
#28	GPT-5.4 mini gpt-5.4-mini	OpenAI	closed	33.5%
#29	GLM-4.7 glm-4.7	Z.AI	open	33.3%
#30	Nemotron 3 Ultra nemotron-3-ultra-550b	NVIDIA	open	33.2%
#31	MiniMax M2.7 minimax-m2-7	MiniMax	open	32.9%
#32	MiMo-V2.5 mimo-v2.5	Xiaomi	open	32.3%
#33	Muse Spark muse-spark	Meta	closed	32.2%
#34	Qwen3.6 27B qwen3.6-27b	Alibaba Cloud	open	32.0%
#35	Qwen3.6 Plus qwen3-6-plus	Alibaba Cloud	closed	31.8%
#36	GPT-5.4 nano gpt.5.4.nano	OpenAI	closed	30.0%
#37	Grok 4.3 grok-4.3	xAI	closed	29.2%
#38	GPT-5 gpt-5	OpenAI	closed	28.7%
#39	Claude Sonnet 4.5 claude-sonnet-4-5	Anthropic	closed	27.6%
#40	Kimi K2.5 kimi-k2.5	Moonshot AI	—	25.4%
#41	GPT-5.1 gpt.5.1	OpenAI	closed	24.4%
#42	Qwen3.5 122B qwen3.5-122b-a10b	Alibaba Cloud	open	23.9%
#43	Gemini 3.1 Pro Preview gemini-3.1-pro-preview	Google	closed	23.1%
#44	Qwen3.5 397B qwen3.5-397b-a17b	Alibaba Cloud	open	23.1%
#45	Qwen3.7-Plus qwen3.7-plus	Alibaba Cloud	—	21.8%
#46	GPT-5 mini gpt-5-mini	OpenAI	closed	21.8%
#47	GLM-4.6 glm-4.6	Z.AI	open	21.7%
#48	Mistral Medium 3.5 mistral-medium-3-5	Mistral AI	open	21.4%
#49	Claude Haiku 4.5 claude-haiku-4-5-20251001	Anthropic	closed	20.3%
#50	DeepSeek V3.1 Terminus deepseek-v3.1-terminus	DeepSeek	open	19.5%
#51	DeepSeek V3.2 deepseek-v3.2	DeepSeek	open	18.3%
#52	Claude Sonnet 4 claude-sonnet-4	Anthropic	closed	18.0%
#53	Claude Opus 4.1 claude-opus-4-1	Anthropic	closed	15.2%
#54	Gemma 4 31B gemma-4-31b	Google	open	15.2%
#55	Gemma 4 26B gemma-4-26b-a4b	Google	open	13.1%
#56	Command A+ command-a-plus	Cohere	open	10.7%
#57	Nemotron 3 Super nemotron-3-super-120b	NVIDIA	open	9.7%
#58	Nova 2 Pro nova-2-pro-preview	Amazon	closed	8.8%
#59	Gemini 2.5 Pro gemini-2.5-pro	Google	closed	8.3%
#60	Gemini 3.1 Flash-Lite gemini-3.1-flash-lite	Google	closed	7.1%
#61	Mistral Large 3 mistral-large-2512	Mistral AI	open	6.6%
#62	GLM-5V-Turbo glm-5v-turbo	Z.AI	closed	4.5%
#63	Mistral Small 4 mistral-small-2603	Mistral AI	open	4.4%
#64	Grok Code Fast 1 grok-code-fast-1	xAI	closed	2.7%
#65	Ling 2.6 1T ling-2.6-1t	InclusionAI	open	0%
#66	MiMo-V2-Omni mimo-v2-omni	Xiaomi	closed	0%

The top of this snapshot is led by Claude Fable 5 at 63.0%; third place is 3.8 points behind. The top-10 spread is 12.3 points.

Compare top models

Claude Fable 5 vs GPT-5.6 Sol GPT-5.6 Sol vs Kimi K3 Kimi K3 vs Claude Opus 4 Claude Opus 4 vs Claude Sonnet 5

About GDPval-AA v2

Definition and scoring

Organisation: Artificial Analysis / OpenAI
Category: Agents
Version: v2
Direction: higher is better
Ranking use: Weighted 1.7%
Contamination risk: Low

Artificial Analysis agentic evaluation of OpenAI GDPval real-world economic work tasks across occupations, scored with pairwise Elo then normalized. Results stay tied to the exact model variant and evaluation system. Multiple systems for the same model use the best published score on this page; overall Lumina scoring uses the median of ranking-eligible rows.

Open benchmark source ↗

FAQ

Quick answers

What does GDPval-AA v2 measure?

Artificial Analysis agentic evaluation of OpenAI GDPval real-world economic work tasks across occupations, scored with pairwise Elo then normalized.

Which model scores highest on GDPval-AA v2?

Claude Fable 5 by Anthropic currently leads with 63.0%.

How many models are evaluated on GDPval-AA v2?

66 models in the LuminaBench cohort have a qualifying score on this benchmark.

Does this affect overall Lumina rank?

Yes. This family is ranking-weighted at 1.7% of its capability category. Claude Fable 5 is overall #2.

Terminal-Bench

2.1 · 80 results · Weighted 1.6%

Berkeley Function-Calling Leaderboard

V3 · 2 results · Not Weighted

τ-bench

Rolling · 37 results · Weighted 1.6%

GAIA

Rolling · 13 results · Weighted 1.6%

BrowseComp

Rolling · 24 results · Weighted 1.6%

OSWorld

Rolling · 11 results · Weighted 1.6%

Agents · Benchmark profile

GDPval-AA v2

Artificial Analysis agentic evaluation of OpenAI GDPval real-world economic work tasks across occupations, scored with pairwise Elo then normalized.

Ranking weightedv2Active66 models

Data verified 15 July 2026 · Methodology 1.2.0

Benchmark score on GDPval-AA v2

Leaderboard · 66 models

Claude Fable 5 leads at 63.0%, followed by GPT-5.6 Sol (62.4%) and Kimi K3 (59.2%).

Ranking · 1.7%Agentsv2

#1Closed

Claude Fable 5

Anthropic

claude-fable-5

63.0%

Overall #2 · Context 1M tokens

#2Closed

GPT-5.6 Sol

OpenAI

gpt-5.6-sol

62.4%

Overall #4 · Context 1.05M tokens

#3Closed

Kimi K3

Moonshot AI

kimi-k3

59.2%

Overall #3 · Context 1M tokens

Visual analysis

See the field, not just the table.

Three views of the same sourced leaderboard: model placement, the score curve, and descriptive provider averages. These visuals never alter ranking calculations.

ModelsColours

AnthropicOpenAIMoonshot AIxAIZ.AIMiniMaxMetaGoogleDeepSeekXiaomiAlibaba Cloud

Leaderboard bars

20 ranked models · higher is better

View accessible chart data

Accessible data for Leaderboard bars
Model	Rank	Provider	Score
Claude Fable 5	#1	Anthropic	63.0%
GPT-5.6 Sol	#2	OpenAI	62.4%
Kimi K3	#3	Moonshot AI	59.2%
Claude Opus 4	#4	Anthropic	56.5%
Claude Sonnet 5	#5	Anthropic	55.3%
Claude Opus 4.8	#6	Anthropic	55.0%
GPT-5.6 Terra	#7	OpenAI	54.6%
GPT-5.6 Luna	#8	OpenAI	54.6%
Grok 4.5	#9	xAI	51.7%
GLM-5.2	#10	Z.AI	50.7%
Claude Opus 4.7	#11	Anthropic	50.0%
GPT-5.5	#12	OpenAI	49.7%
MiniMax M3	#13	MiniMax	44.7%
GPT-5.4	#14	OpenAI	44.7%
Claude Sonnet 4.6	#15	Anthropic	43.9%
Muse Spark 1.1	#16	Meta	43.7%
Gemini 3.5 Flash	#17	Google	42.4%
DeepSeek V4 Pro	#18	DeepSeek	40.4%
MiMo-V2-Flash	#19	Xiaomi	40.4%
Qwen3.7-Max	#20	Alibaba Cloud	38.7%

Score curve

Shows how quickly scores separate down the field

View accessible chart data

Accessible data for Score curve
Model	Rank	Provider	Score
Claude Fable 5	#1	Anthropic	63.0%
GPT-5.6 Sol	#2	OpenAI	62.4%
Kimi K3	#3	Moonshot AI	59.2%
Claude Opus 4	#4	Anthropic	56.5%
Claude Sonnet 5	#5	Anthropic	55.3%
Claude Opus 4.8	#6	Anthropic	55.0%
GPT-5.6 Terra	#7	OpenAI	54.6%
GPT-5.6 Luna	#8	OpenAI	54.6%
Grok 4.5	#9	xAI	51.7%
GLM-5.2	#10	Z.AI	50.7%
Claude Opus 4.7	#11	Anthropic	50.0%
GPT-5.5	#12	OpenAI	49.7%
MiniMax M3	#13	MiniMax	44.7%
GPT-5.4	#14	OpenAI	44.7%
Claude Sonnet 4.6	#15	Anthropic	43.9%
Muse Spark 1.1	#16	Meta	43.7%
Gemini 3.5 Flash	#17	Google	42.4%
DeepSeek V4 Pro	#18	DeepSeek	40.4%
MiMo-V2-Flash	#19	Xiaomi	40.4%
Qwen3.7-Max	#20	Alibaba Cloud	38.7%

Provider profile

Average within the visible cohort · descriptive only

View accessible chart data

Accessible data for Provider profile
Provider	Models	Average score
Moonshot AI	1	59.2%
Anthropic	6	53.9%
OpenAI	5	53.2%
xAI	1	51.7%
Z.AI	1	50.7%
MiniMax	1	44.7%
Meta	1	43.7%
Google	1	42.4%
DeepSeek	1	40.4%
Xiaomi	1	40.4%
Alibaba Cloud	1	38.7%

Benchmark score table (66 models)

One best score per model · higher is better

GDPval-AA v2 leaderboard ranked by score
Rank	Model	Provider	License	Score
#1	Claude Fable 5 claude-fable-5	Anthropic	closed	63.0%
#2	GPT-5.6 Sol gpt-5.6-sol	OpenAI	closed	62.4%
#3	Kimi K3 kimi-k3	Moonshot AI	closed	59.2%
#4	Claude Opus 4 claude-opus-4	Anthropic	closed	56.5%
#5	Claude Sonnet 5 claude-sonnet-5	Anthropic	closed	55.3%
#6	Claude Opus 4.8 claude-opus-4-8	Anthropic	closed	55.0%
#7	GPT-5.6 Terra gpt-5.6-terra	OpenAI	closed	54.6%
#8	GPT-5.6 Luna gpt-5.6-luna	OpenAI	closed	54.6%
#9	Grok 4.5 grok-4.5	xAI	closed	51.7%
#10	GLM-5.2 glm-5-2	Z.AI	open	50.7%
#11	Claude Opus 4.7 claude-opus-4-7	Anthropic	closed	50.0%
#12	GPT-5.5 gpt-5.5	OpenAI	closed	49.7%
#13	MiniMax M3 MiniMax-M3	MiniMax	—	44.7%
#14	GPT-5.4 gpt-5.4	OpenAI	closed	44.7%
#15	Claude Sonnet 4.6 claude-sonnet-4-6	Anthropic	closed	43.9%
#16	Muse Spark 1.1 muse-spark-1-1	Meta	closed	43.7%
#17	Gemini 3.5 Flash gemini-3.5-flash	Google	closed	42.4%
#18	DeepSeek V4 Pro deepseek-v4-pro	DeepSeek	open	40.4%
#19	MiMo-V2-Flash mimo-v2-flash	Xiaomi	open	40.4%
#20	Qwen3.7-Max qwen3.7-max	Alibaba Cloud	—	38.7%
#21	MiMo-V2.5-Pro mimo-v2-5-pro	Xiaomi	closed	38.3%
#22	GLM-5.1 glm-5-1	Z.AI	open	37.8%
#23	Claude Sonnet 3.7 claude-3-7-sonnet	Anthropic	closed	35.9%
#24	Hy3 hy3	Tencent	open	35.7%
#25	Kimi K2.6 kimi-k2.6	Moonshot AI	open	34.5%
#26	DeepSeek V4 Flash deepseek-v4-flash	DeepSeek	open	34.4%
#27	Kimi K2.7 Code kimi-k2.7-code	Moonshot AI	open	34.3%
#28	GPT-5.4 mini gpt-5.4-mini	OpenAI	closed	33.5%
#29	GLM-4.7 glm-4.7	Z.AI	open	33.3%
#30	Nemotron 3 Ultra nemotron-3-ultra-550b	NVIDIA	open	33.2%
#31	MiniMax M2.7 minimax-m2-7	MiniMax	open	32.9%
#32	MiMo-V2.5 mimo-v2.5	Xiaomi	open	32.3%
#33	Muse Spark muse-spark	Meta	closed	32.2%
#34	Qwen3.6 27B qwen3.6-27b	Alibaba Cloud	open	32.0%
#35	Qwen3.6 Plus qwen3-6-plus	Alibaba Cloud	closed	31.8%
#36	GPT-5.4 nano gpt.5.4.nano	OpenAI	closed	30.0%
#37	Grok 4.3 grok-4.3	xAI	closed	29.2%
#38	GPT-5 gpt-5	OpenAI	closed	28.7%
#39	Claude Sonnet 4.5 claude-sonnet-4-5	Anthropic	closed	27.6%
#40	Kimi K2.5 kimi-k2.5	Moonshot AI	—	25.4%
#41	GPT-5.1 gpt.5.1	OpenAI	closed	24.4%
#42	Qwen3.5 122B qwen3.5-122b-a10b	Alibaba Cloud	open	23.9%
#43	Gemini 3.1 Pro Preview gemini-3.1-pro-preview	Google	closed	23.1%
#44	Qwen3.5 397B qwen3.5-397b-a17b	Alibaba Cloud	open	23.1%
#45	Qwen3.7-Plus qwen3.7-plus	Alibaba Cloud	—	21.8%
#46	GPT-5 mini gpt-5-mini	OpenAI	closed	21.8%
#47	GLM-4.6 glm-4.6	Z.AI	open	21.7%
#48	Mistral Medium 3.5 mistral-medium-3-5	Mistral AI	open	21.4%
#49	Claude Haiku 4.5 claude-haiku-4-5-20251001	Anthropic	closed	20.3%
#50	DeepSeek V3.1 Terminus deepseek-v3.1-terminus	DeepSeek	open	19.5%
#51	DeepSeek V3.2 deepseek-v3.2	DeepSeek	open	18.3%
#52	Claude Sonnet 4 claude-sonnet-4	Anthropic	closed	18.0%
#53	Claude Opus 4.1 claude-opus-4-1	Anthropic	closed	15.2%
#54	Gemma 4 31B gemma-4-31b	Google	open	15.2%
#55	Gemma 4 26B gemma-4-26b-a4b	Google	open	13.1%
#56	Command A+ command-a-plus	Cohere	open	10.7%
#57	Nemotron 3 Super nemotron-3-super-120b	NVIDIA	open	9.7%
#58	Nova 2 Pro nova-2-pro-preview	Amazon	closed	8.8%
#59	Gemini 2.5 Pro gemini-2.5-pro	Google	closed	8.3%
#60	Gemini 3.1 Flash-Lite gemini-3.1-flash-lite	Google	closed	7.1%
#61	Mistral Large 3 mistral-large-2512	Mistral AI	open	6.6%
#62	GLM-5V-Turbo glm-5v-turbo	Z.AI	closed	4.5%
#63	Mistral Small 4 mistral-small-2603	Mistral AI	open	4.4%
#64	Grok Code Fast 1 grok-code-fast-1	xAI	closed	2.7%
#65	Ling 2.6 1T ling-2.6-1t	InclusionAI	open	0%
#66	MiMo-V2-Omni mimo-v2-omni	Xiaomi	closed	0%

The top of this snapshot is led by Claude Fable 5 at 63.0%; third place is 3.8 points behind. The top-10 spread is 12.3 points.

Compare top models

Claude Fable 5 vs GPT-5.6 Sol GPT-5.6 Sol vs Kimi K3 Kimi K3 vs Claude Opus 4 Claude Opus 4 vs Claude Sonnet 5

About GDPval-AA v2

Definition and scoring

Organisation: Artificial Analysis / OpenAI
Category: Agents
Version: v2
Direction: higher is better
Ranking use: Weighted 1.7%
Contamination risk: Low

Open benchmark source ↗

FAQ

Quick answers

What does GDPval-AA v2 measure?

Artificial Analysis agentic evaluation of OpenAI GDPval real-world economic work tasks across occupations, scored with pairwise Elo then normalized.

Which model scores highest on GDPval-AA v2?

Claude Fable 5 by Anthropic currently leads with 63.0%.

How many models are evaluated on GDPval-AA v2?

66 models in the LuminaBench cohort have a qualifying score on this benchmark.

Does this affect overall Lumina rank?

Yes. This family is ranking-weighted at 1.7% of its capability category. Claude Fable 5 is overall #2.

Terminal-Bench

2.1 · 80 results · Weighted 1.6%

Berkeley Function-Calling Leaderboard

V3 · 2 results · Not Weighted

τ-bench

Rolling · 37 results · Weighted 1.6%

GAIA

Rolling · 13 results · Weighted 1.6%

BrowseComp

Rolling · 24 results · Weighted 1.6%

OSWorld

Rolling · 11 results · Weighted 1.6%

Leaderboard · 66 models

Claude Fable 5

GPT-5.6 Sol

Kimi K3

See the field, not just the table.

Benchmark score table (66 models)

Compare top models

Definition and scoring

Quick answers

What does GDPval-AA v2 measure?

Which model scores highest on GDPval-AA v2?

How many models are evaluated on GDPval-AA v2?

Does this affect overall Lumina rank?

More Agents benchmarks

Leaderboard · 66 models

Claude Fable 5

GPT-5.6 Sol

Kimi K3

See the field, not just the table.

Benchmark score table (66 models)

Compare top models

Definition and scoring

Quick answers

What does GDPval-AA v2 measure?

Which model scores highest on GDPval-AA v2?

How many models are evaluated on GDPval-AA v2?

Does this affect overall Lumina rank?

More Agents benchmarks