Agents · Benchmark profile

GAIA

A benchmark for general assistants that must reason, browse, use tools and combine modalities.

Ranking weightedCurrent / rollingActive12 models

Data verified 15 Jul 2026 · Methodology 2.3.0

Benchmark score on GAIA

Leaderboard · 12 models

Claude Fable 5 leads at 52.3%, followed by Claude Mythos 5 (52.3%) and GPT-5.4 (48.2%).

Ranking · 1.6%AgentsCurrent

#1Closed

Claude Fable 5

Anthropic

claude-fable-5

52.3%

Overall #2 · Context 1M tokens

#2Closed

Claude Mythos 5

Anthropic

claude-mythos-5

52.3%

Overall — · Context 1M tokens

#3Closed

GPT-5.4

OpenAI

gpt-5.4

48.2%

Overall #19 · Context 1.05M tokens

ViewModelsColours

AnthropicOpenAIGooglexAIAlibaba CloudMoonshot AI

Lumina

GAIA leaderboard

Accessible data for GAIA leaderboard
Model	Rank	Model ID	Provider	Score	Unit
Claude Fable 5	#1	claude-fable-5	Anthropic	52.3	percent
Claude Mythos 5	#2	claude-mythos-5	Anthropic	52.3	percent
GPT-5.4	#3	gpt-5-4	OpenAI	48.2	percent
Claude Opus 4.6	#4	claude-opus-4-6	Anthropic	47.8	percent
Gemini 3.1 Pro Preview	#5	gemini-3-1-pro-preview	Google	46.1	percent
Claude Sonnet 4.6	#6	claude-sonnet-4-6	Anthropic	45.5	percent
Grok 4.1	#7	grok-4-1	xAI	39.7	percent
Qwen3.6 Plus	#8	qwen3-6-plus	Alibaba Cloud	37.4	percent
Kimi K2.5	#9	kimi-k2-5	Moonshot AI	36.5	percent
Gemini 3 Flash	#10	gemini-3-flash	Google	35.2	percent
GLM-5	#11	glm-5	Z.AI	31.5	percent
Mistral Large 3	#12	mistral-large-3	Mistral AI	30.8	percent

Benchmark score table (12 models)

One best compatible score per canonical product · higher is better

GAIA leaderboard ranked by score
Rank	Model	Provider	License	Evidence use	Score
#1	Claude Fable 5 claude-fable-5	Anthropic	closed	Source carrier	52.3%
#2	Claude Mythos 5 claude-mythos-5	Anthropic	closed	Source carrier	52.3%
#3	GPT-5.4 gpt-5.4	OpenAI	closed	Source carrier	48.2%
#4	Claude Opus 4.6 claude-opus-4-6	Anthropic	closed	Source carrier	47.8%
#5	Gemini 3.1 Pro Preview gemini-3.1-pro-preview	Google	closed	Source carrier	46.1%
#6	Claude Sonnet 4.6 claude-sonnet-4-6	Anthropic	closed	Source carrier	45.5%
#7	Grok 4.1 grok-4-1	xAI	closed	Source carrier	39.7%
#8	Qwen3.6 Plus qwen3-6-plus	Alibaba Cloud	closed	Source carrier	37.4%
#9	Kimi K2.5 kimi-k2.5	Moonshot AI	open	Source carrier	36.5%
#10	Gemini 3 Flash gemini-3-flash	Google	closed	Source carrier	35.2%
#11	GLM-5 glm-5	Z.AI	open	Source carrier	31.5%
#12	Mistral Large 3 mistral-large-2512	Mistral AI	open	Source carrier	30.8%

The top of this snapshot is led by Claude Fable 5 at 52.3%; third place is 4.1 points behind. The top-10 spread is 17.1 points.

Configuration-preserving evidencePublished source rows (12)Exact source labels, configurations, versions and dated observations. Same-snapshot canonical/configuration projections are shown once.

Every distinct dated source-published row for GAIA, including exact model configuration labels
Exact published model / configuration	Canonical product	Score	Version / system	Evidence	Source / observed
Exact model variant as published on the source leaderboard page. source label without registered configuration ID	Claude Mythos 5	52.3%	CurrentPublic leaderboard extraction	Source carrier · no additional scoring weight source-checked	GAIA leaderboard via BenchLM ↗Observed 2026-07-15 · checked 2026-07-15
Exact model variant as published on the source leaderboard page. source label without registered configuration ID	Claude Fable 5	52.3%	CurrentPublic leaderboard extraction	Source carrier · no additional scoring weight source-checked	GAIA leaderboard via BenchLM ↗Observed 2026-07-15 · checked 2026-07-15
Exact model variant as published on the source leaderboard page. source label without registered configuration ID	GPT-5.4	48.2%	CurrentPublic leaderboard extraction	Source carrier · no additional scoring weight source-checked	GAIA leaderboard via BenchLM ↗Observed 2026-07-15 · checked 2026-07-15
Exact model variant as published on the source leaderboard page. source label without registered configuration ID	Claude Opus 4.6	47.8%	CurrentPublic leaderboard extraction	Source carrier · no additional scoring weight source-checked	GAIA leaderboard via BenchLM ↗Observed 2026-07-15 · checked 2026-07-15
Exact model variant as published on the source leaderboard page. source label without registered configuration ID	Gemini 3.1 Pro Preview	46.1%	CurrentPublic leaderboard extraction	Source carrier · no additional scoring weight source-checked	GAIA leaderboard via BenchLM ↗Observed 2026-07-15 · checked 2026-07-15
Exact model variant as published on the source leaderboard page. source label without registered configuration ID	Claude Sonnet 4.6	45.5%	CurrentPublic leaderboard extraction	Source carrier · no additional scoring weight source-checked	GAIA leaderboard via BenchLM ↗Observed 2026-07-15 · checked 2026-07-15
Exact model variant as published on the source leaderboard page. source label without registered configuration ID	Grok 4.1	39.7%	CurrentPublic leaderboard extraction	Source carrier · no additional scoring weight source-checked	GAIA leaderboard via BenchLM ↗Observed 2026-07-15 · checked 2026-07-15
Exact model variant as published on the source leaderboard page. source label without registered configuration ID	Qwen3.6 Plus	37.4%	CurrentPublic leaderboard extraction	Source carrier · no additional scoring weight source-checked	GAIA leaderboard via BenchLM ↗Observed 2026-07-15 · checked 2026-07-15
Exact model variant as published on the source leaderboard page. source label without registered configuration ID	Kimi K2.5	36.5%	CurrentPublic leaderboard extraction	Source carrier · no additional scoring weight source-checked	GAIA leaderboard via BenchLM ↗Observed 2026-07-15 · checked 2026-07-15
Exact model variant as published on the source leaderboard page. source label without registered configuration ID	Gemini 3 Flash	35.2%	CurrentPublic leaderboard extraction	Source carrier · no additional scoring weight source-checked	GAIA leaderboard via BenchLM ↗Observed 2026-07-15 · checked 2026-07-15
Exact model variant as published on the source leaderboard page. source label without registered configuration ID	GLM-5	31.5%	CurrentPublic leaderboard extraction	Source carrier · no additional scoring weight source-checked	GAIA leaderboard via BenchLM ↗Observed 2026-07-15 · checked 2026-07-15
Exact model variant as published on the source leaderboard page. source label without registered configuration ID	Mistral Large 3	30.8%	CurrentPublic leaderboard extraction	Source carrier · no additional scoring weight source-checked	GAIA leaderboard via BenchLM ↗Observed 2026-07-15 · checked 2026-07-15

Compare top models

Claude Fable 5 vs Claude Mythos 5 Claude Mythos 5 vs GPT-5.4 GPT-5.4 vs Claude Opus 4.6 Claude Opus 4.6 vs Gemini 3.1 Pro Preview

About GAIA

Definition and scoring

Organisation: GAIA authors
Category: Agents
Version: Not evaluated
Direction: higher is better
Ranking use: Weighted 1.6%
Contamination risk: Unknown

A benchmark for general assistants that must reason, browse, use tools and combine modalities. Every genuine source row stays tied to its exact model label, configuration, benchmark version and evaluation system. The summary chart shows one best compatible score per canonical product; Score 2.0 admits only explicitly mapped, frozen protocols and keeps incompatible configurations separate.

Open benchmark source ↗

FAQ

Quick answers

What does GAIA measure?

A benchmark for general assistants that must reason, browse, use tools and combine modalities.

Which model scores highest on GAIA?

Claude Fable 5 by Anthropic currently leads with 52.3%.

How many models are evaluated on GAIA?

12 canonical model products have a qualifying score, backed by 12 distinct dated source rows.

Does this affect overall Lumina rank?

Yes. This family is ranking-weighted at 1.6% of its capability category. Claude Fable 5 is overall #2.

Agents · Benchmark profile

GAIA

A benchmark for general assistants that must reason, browse, use tools and combine modalities.

Ranking weightedCurrent / rollingActive12 models

Data verified 15 Jul 2026 · Methodology 2.3.0

Benchmark score on GAIA