GPQA Diamond

AI benchmark results and model performance

Type: percentageGeneral Knowledge

View benchmark source

Recorded Results

185

Average Score

67.50%

Scores Over Time

Individual benchmark scores plotted by date.

Models Using This Benchmark

Organisation	Model	Reported	Top Score	Info	Self Reported	Source
Google	Gemma 3 1B	12 Mar 2025	19.20%	-	Yes	Source
Qwen	Qwen 2 7B Instruct	-	25.30%	-	Yes	Source
Google	Gemini 1.0 Pro	06 Dec 2023	27.90%	-	No	-
ai21	Jamba Mini 1.6	06 Mar 2025	30%	-	No	Source
Meta	Llama 3.1 8B Instruct	23 Jul 2024	30.40%	-	Yes	Source
Microsoft	Phi 3.5 mini instruct	23 Aug 2024	30.40%	-	Yes	Source
Google	Gemma 3 4B	12 Mar 2025	30.80%	-	Yes	Source
OpenAI	GPT 3.5 Turbo 0613	13 Jun 2023	30.80%	-	No	-
Qwen	Qwen 2.5 Omni 7B	-	30.80%	-	Yes	Source
ai21	Jamba Mini 1.5	22 Aug 2024	32.30%	-	Yes	Source
Meta	Llama 3.2 3B Instruct	25 Sept 2024	32.80%	-	Yes	Source
Anthropic	Claude 3 Haiku	13 Mar 2024	33.30%	-	Yes	Source
Google	Gemini 1.0 Ultra	06 Dec 2023	35.70%	-	Yes	Source
OpenAI	GPT 4 (2023-03-14)	14 Mar 2023	35.70%	-	Yes	Source
SpaceXAI	Grok 1.5	28 Mar 2024	35.90%	-	Yes	Source
Qwen	Qwen 2.5 7B	-	36.40%	-	Yes	Source
Microsoft	Phi 3.5 MoE instruct	23 Aug 2024	36.80%	-	Yes	Source
ai21	Jamba Large 1.5	22 Aug 2024	36.90%	-	Yes	Source
ai21	Jamba Large 1.6	06 Mar 2025	38.70%	-	No	Source
Amazon	Nova Micro 1.0	04 Dec 2024	40%	-	Yes	Source
OpenAI	GPT 4o Mini (2024-07-18)	18 Jul 2024	40.20%	-	Yes	Source
Anthropic	Claude 3 Sonnet	04 Mar 2024	40.40%	-	Yes	Source
Google	Gemini Diffusion	20 May 2025	40.40%	Pass@1	Yes	Source
Google	Gemma 3 12B	12 Mar 2025	40.90%	-	Yes	Source
Anthropic	Claude 3.5 Haiku	04 Nov 2024	41.60%	-	Yes	Source
Meta	Llama 3.1 70B Instruct	23 Jul 2024	41.70%	-	Yes	Source
Amazon	Nova Lite 1.0	04 Dec 2024	42%	-	Yes	Source
Google	Gemma 3 27B	12 Mar 2025	42.40%	-	Yes	Source
Qwen	Qwen 2 72B Instruct	-	42.40%	-	Yes	Source
Mistral	Mistral Large 3.0	02 Dec 2025	43.90%	5 Shot, No Reasoning	Yes	Source
Mistral	Mistral Small 3.0	30 Jan 2025	45.30%	-	Yes	Source
Qwen	Qwen 2.5 14B	-	45.50%	-	Yes	Source
Mistral	Mistral Small 3.1	17 Mar 2025	46%	-	Yes	Source
OpenAI	GPT 4o (2024-08-06)	06 Aug 2024	46%	-	Yes	Source
Qwen	Qwen 2.5 VL 32B Instruct	24 Mar 2025	46%	-	Yes	Source
Mistral	Mistral Small 3.2	20 Jun 2025	46.13%	-	Yes	Source
Amazon	Nova Pro 1.0	04 Dec 2024	46.90%	-	Yes	Source
Qwen	Qwen 3 235B A22B	-	47.50%	-	Yes	Source
Moonshot	Kimi K2 (2025-07-11)	11 Jul 2025	48.10%	Avg@8	Yes	Source
Qwen	Qwen 2.5 72B	-	49%	-	Yes	Source
Qwen	Qwen 2.5 32B	-	49.50%	-	Yes	Source
OpenAI	GPT 4.1 Nano	14 Apr 2025	50.30%	-	Yes	Source
Anthropic	Claude 3 Opus	04 Mar 2024	50.40%	-	Yes	Source
Meta	Llama 3.3 70B Instruct	06 Dec 2024	50.50%	-	Yes	Source
Meta	Llama 3.1 405B Instruct	23 Jul 2024	50.70%	-	Yes	Source
SpaceXAI	Grok 2 Mini	13 Aug 2024	51%	-	Yes	Source
LG	EXAONE 4.0 1.2B	15 Jul 2025	52%	Reasoning	Yes	Source
Microsoft	Phi 4 Mini Reasoning	30 Apr 2025	52%	-	Yes	Source
Mistral	Ministral 3.0 3B	02 Dec 2025	53.40%	-	Yes	Source
Nvidia	Llama 3.1 Nemotron Nano 8B V1	18 Mar 2025	54.10%	-	Yes	Source
SpaceXAI	Grok 2	13 Aug 2024	56%	-	Yes	Source
Microsoft	Phi 4	12 Dec 2024	56.10%	-	Yes	Source
Amazon	Nova Premier	30 Apr 2025	57.10%	-	Yes	-
DeepSeek	DeepSeek V3 (2024-12-26)	26 Dec 2024	59.10%	-	No	Source
OpenAI	o1 mini	12 Sept 2024	60%	-	Yes	Source
Anthropic	Claude 3.5 Sonnet (2024-10-22)	22 Oct 2024	65%	-	Yes	-
OpenAI	GPT 4.1 Mini	14 Apr 2025	65%	-	Yes	Source
Qwen	QwQ 32B	-	65.20%	-	Yes	Source
Qwen	QwQ 32B Preview	-	65.20%	-	Yes	Source
Microsoft	Phi 4 Reasoning	30 Apr 2025	65.80%	-	Yes	Source
Qwen	Qwen 3 30B A3B	-	65.80%	-	Yes	Source
OpenAI	GPT 4.1	14 Apr 2025	66.30%	-	Yes	Source
Google	Gemini 2.5 Flash Lite Preview (2025-06-17)	17 Jun 2025	66.70%	Thinking	Yes	Source
Nvidia	Llama 3.3 Nemotron Super 49B v1	18 Mar 2025	66.70%	-	Yes	Source
Mistral	Ministral 3.0 8B	02 Dec 2025	66.80%	-	Yes	Source
Anthropic	Claude 3.5 Sonnet (2024-06-20)	21 Jun 2024	67.20%	-	Yes	Source
Mistral	Magistral Small 1.0	10 Jun 2025	68.20%	-	Yes	Source
DeepSeek	DeepSeek V3 (2025-03-24)	24 Mar 2025	68.40%	-	Yes	Source
Microsoft	Phi 4 Reasoning Plus	30 Apr 2025	68.90%	-	Yes	Source
Mistral	Magistral Medium 1.0	10 Jun 2025	70.80%	-	Yes	Source
Mistral	Ministral 3.0 14B	02 Dec 2025	71.20%	-	Yes	Source
OpenAI	GPT 5 Nano	07 Aug 2025	71.20%	High Reasoning Effort, No Tools	Yes	Source
OpenAI	GPT 4.5 Preview	27 Feb 2025	71.40%	-	Yes	-
DeepSeek	DeepSeek R1 (2025-01-20)	20 Jan 2025	71.50%	-	No	Source
Upstage	Solar Pro 3 (2026-03-23)	23 Mar 2026	72.40%	Artificial Analysis structured model metrics	No	Source
Google	DiffusionGemma 26B A4B IT	10 Jun 2026	73.20%	-	Yes	Source
OpenAI	o1 preview	12 Sept 2024	73.30%	-	Yes	Source
OpenAI	GPT OSS 20b	05 Aug 2025	74.20%	High Reasoning Effort, With Tools	Yes	Source
Moonshot	Kimi K2 (2025-09-05)	05 Sept 2025	75.10%	Avg@8	Yes	Source
LG	EXAONE 4.0 32B	15 Jul 2025	75.40%	Reasoning	Yes	Source
Nvidia	Nemotron 3.5 Lightning 30B A3B	11 Aug 2026	75.44%	No tools; NVIDIA release evaluation suite	Yes	Source
Nvidia	Llama 3.1 Nemotron Ultra 253B v1	07 Apr 2025	76%	-	Yes	Source
Arcee AI	Trinity Large Thinking	01 Apr 2026	76.30%	Hugging Face model card benchmark table (arcee-ai/Trinity-Large-Thinking)	Yes	Source
Qwen	Qwen 3 A235 A22B Instruct 2507	-	77.50%	-	Yes	Source
OpenAI	o1	17 Dec 2024	78%	-	Yes	Source
Google	Gemini 2.0 Flash	05 Feb 2025	78.30%	Single Attempt	Yes	Source
Google	Gemini 2.5 Flash Preview (2025-04-17)	17 Apr 2025	78.30%	Pass@1	Yes	Source
Google	Gemma 4 12B	03 Jun 2026	78.80%	-	Yes	Source
ByteDance	Seed 2.0 Mini	14 Feb 2026	79%	Seed2 official benchmark table \| GPQA Diamond	Yes	Source
OpenAI	o1 pro	19 Mar 2025	79%	-	Yes	Source
SpaceXAI	Grok 3	18 Apr 2025	79.10%	-	Yes	Source
Amazon	Nova 2 Lite	02 Dec 2025	79.60%	-	Yes	Source
OpenAI	o3 mini	30 Jan 2025	79.70%	High Reasoning Effort	Yes	-
SpaceXAI	Grok 3 Mini	18 Apr 2025	80.30%	High Reasoning Effort	Yes	Source
Anthropic	Claude Opus 4.1	05 Aug 2025	80.90%	-	Yes	Source
OpenAI	GPT OSS 120b	05 Aug 2025	80.90%	High Reasoning Effort, With Tools	Yes	Source
DeepSeek	DeepSeek R1 (2025-05-28)	28 May 2025	81%	-	Yes	Source
Qwen	Qwen 3 235B A22B Thinking 2507	-	81.10%	-	Yes	Source
Amazon	Nova 2 Pro	02 Dec 2025	81.40%	-	Yes	Source
OpenAI	o4 Mini	16 Apr 2025	81.40%	-	Yes	Source
Google	Gemma 4 26B A4B	02 Apr 2026	82.30%	-	Yes	Source
OpenAI	GPT 5 Mini	07 Aug 2025	82.30%	High Reasoning Effort, No Tools	Yes	Source
Google	Gemini 2.5 Flash Preview (2025-05-20)	20 May 2025	82.80%	Pass@1	Yes	Source
OpenAI	GPT 5.4 Nano	17 Mar 2026	82.80%	-	Yes	Source
Google	Gemini 2.5 Pro Preview (2025-05-06)	06 May 2025	83%	Pass@1	Yes	Source
Anthropic	Claude Opus 4	21 May 2025	83.30%	-	Yes	-
OpenAI	o3	16 Apr 2025	83.30%	-	Yes	Source
Meta	Muse Glimmer 30B	10 Aug 2026	83.50%	AA; high reasoning	Yes	Source
Anthropic	Claude Sonnet 4	21 May 2025	83.80%	-	Yes	-
OpenAI	o3 Pro	10 Jun 2025	84%	-	Yes	Source
SpaceXAI	Grok 3 Mini Beta	19 Feb 2025	84%	Reasoning	Yes	Source
Google	Gemma 4 31B	02 Apr 2026	84.30%	-	Yes	Source
SpaceXAI	Grok 3 Beta	19 Feb 2025	84.60%	Think, Cons@64	Yes	Source
z.AI	GLM 5 Turbo	15 Mar 2026	84.70%	Artificial Analysis structured model metrics	No	Source
Anthropic	Claude 3.7 Sonnet	24 Feb 2025	84.80%	-	Yes	Source
ByteDance	Seed 2.0 Lite	14 Feb 2026	85.10%	Seed2 official benchmark table \| GPQA Diamond	Yes	Source
DeepSeek	DeepSeek V3.2 Speciale	01 Dec 2025	85.70%	-	Yes	Source
z.AI	GLM 5.1	07 Apr 2026	86.20%	-	Yes	Source
Google	Gemini 2.5 Pro Preview (2025-06-05)	05 Jun 2025	86.40%	Single Attempt	Yes	Source
Google	Gemini 3.1 Flash-Lite	07 May 2026	86.90%	-	Yes	Source
Google	Gemini 3.1 Flash Lite Preview	03 Mar 2026	86.90%	-	Yes	Source
Anthropic	Claude Opus 4.5	24 Nov 2025	86.95%	Avg@5, 64k Thinking	Yes	Source
OpenAI	GPT 5	07 Aug 2025	87.30%	Pass @ 1	Yes	Source
SpaceXAI	Grok 4	10 Jul 2025	87.50%	No Tools	Yes	Source
OpenAI	o3 Preview	20 Dec 2024	87.70%	-	Yes	Source
OpenAI	GPT 5.4 Mini	17 Mar 2026	88%	-	Yes	Source
DeepSeek	DeepSeek V4 Flash	24 Apr 2026	88.10%	Pass@1; DeepSeek-V4-Flash Max	Yes	Source
ByteDance	Seed 2.0 Pro	14 Feb 2026	88.90%	Seed2 official benchmark table \| GPQA Diamond	Yes	Source
SpaceXAI	Grok 4 Heavy	10 Jul 2025	88.90%	-	Yes	Source
Qwen	Qwen 3.8 27B	14 Aug 2026	89.20%	-	Yes	Source
Meta	Muse Spark	08 Apr 2026	89.50%	-	Yes	Source
Anthropic	Claude Sonnet 4.6	17 Feb 2026	89.90%	-	Yes	Source
DeepSeek	DeepSeek V4 Pro	24 Apr 2026	90.10%	Pass@1; DeepSeek-V4-Pro Max	Yes	Source
SpaceXAI	Grok 4.20	17 Feb 2026	91.10%	Artificial Analysis structured model metrics	No	Source
Anthropic	Claude Opus 4.6	05 Feb 2026	91.30%	-	Yes	Source
OpenAI	GPT 5.2	11 Dec 2025	92.40%	No Tools	Yes	Source
Qwen	Qwen 3.7 Max	21 May 2026	92.40%	-	Yes	Source
OpenAI	GPT 5.4	05 Mar 2026	92.80%	-	Yes	Source
OpenAI	GPT 5.2 Pro	11 Dec 2025	93.20%	No Tools	Yes	Source
Moonshot	Kimi K3	16 Jul 2026	93.50%	-	Yes	Source
Anthropic	Claude Opus 4.8	28 May 2026	93.60%	Max effort	Yes	Source
OpenAI	GPT 5.5	23 Apr 2026	93.60%	-	Yes	Source
Google	Gemini 3 Pro Preview	18 Nov 2025	93.80%	Deep Think, Tools Off	Yes	Source
Anthropic	Claude Opus 4.7	16 Apr 2026	94.20%	-	Yes	Source
Google	Gemini 3.1 Pro Preview	19 Feb 2026	94.30%	No Tools	Yes	-
OpenAI	GPT 5.4 Pro	05 Mar 2026	94.40%	-	Yes	Source
OpenAI	GPT 5.5 Pro	23 Apr 2026	94.40%	-	Yes	Source
Anthropic	Claude Mythos Preview	07 Apr 2026	94.60%	-	Yes	Source
Sakana AI	Fugu	22 Jun 2026	95.50%	Sakana reports this benchmark as GPQA-D.	Yes	Source
Sakana AI	Fugu Ultra	22 Jun 2026	95.50%	Sakana reports this benchmark as GPQA-D.	Yes	Source

GPQA Diamond

Recorded Results

Average Score

Models Using This Benchmark

Score Range

Leading Model (lowest score)