Gemini 2.5 Flash Jun 2025

Gemini 2.5 Flash ranks first on FaithJudge.

Google DeepMind Released 17 Jun 2025

Benchmarks with a score

Ranked first, out of five or more models

In the top three

Google DeepMind Jun 2025

Gemini 2.5 Flash

Jun 2025 snapshot Closed weights

140.6 #120 of 271

Epoch Capabilities Index

Terminal-Bench 2.0 17.1% #41 /43
CritPt 1.10% #95 /160
SimpleQA Verified 28.1% #34 /56

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

Benchmark	Score	Rank	Source
FaithJudge	6.26%	#1 of 37	Official leaderboard
RewardBench 2	77.7%	#4 of 42	Official leaderboard
SuperGPQA	58.9%	#12 of 58	Official leaderboard
Global MMLU Lite	88.4%	#4 of 16	Model card
LMArena (Elo)	1,409	#119 of 333	Official leaderboard
FACTS Grounding	85.3%	#6 of 15	Model card
SimpleQA	26.9%	#26 of 59	Model card
Wer wird Millionär? LLM Benchmark	5689.0	#37 of 76	Official leaderboard
ECLeKTic	36.8%	#5 of 10	Model card
MRCR v2	74%	#14 of 27	Model card
FACTS Parametric	30.7%	#27 of 45	Official leaderboard
SimpleQA Verified	28.1%	#34 of 56	Official leaderboard
LOFT	82.1%	#4 of 6	Model card
Humanity's Last Exam	11%	#130 of 147	Model card

Math and science

Benchmark	Score	Rank	Source
Misguided Attention	62.5%	#6 of 45	Official leaderboard
GPQA Diamond	82.8%	#93 of 284	Model card
PHYBench	33.5%	#10 of 28	Official leaderboard
ForecastBench	62.9	#39 of 83	Epoch AI
FrontierMath Tier 4	4.17%	#30 of 55	Epoch AI
DTBench	60.9%	#100 of 178	Official leaderboard
CritPt	1.10%	#95 of 160	Epoch AI
SimpleBench	41.2%	#52 of 82	Epoch AI
FrontierMath	4.84%	#46 of 70	Epoch AI
AIME 2025	72%	#109 of 143	Model card
HiddenMath-Hard	75.5%	#2 of 2	Model card

Coding

Benchmark	Score	Rank	Source
LiveCodeBench v5	63.9%	#18 of 44	Model card
AutoCodeBench	45.7%	#14 of 34	Official leaderboard
Aider Polyglot	61.9%	#13 of 29	Model card
LiveCodeBench	59.3%	#34 of 73	Model card
WeirdML	41%	#74 of 112	Official leaderboard
ALE-Bench	662	#86 of 121	Epoch AI
SWE-bench Verified	28.7%	#49 of 60	Official leaderboard
IOI (Vals)	2.61%	#48 of 55	Official leaderboard

Agents and tools

Benchmark	Score	Rank	Source
Step Race	4	#6 of 59	Official leaderboard
Elimination Game	5	#17 of 56	Official leaderboard
BALROG	33.5%	#13 of 35	Official leaderboard
LisanBench	362.3	#76 of 133	Official leaderboard
DeepSearchQA	25.9%	#6 of 10	Official leaderboard
Galileo Agent Leaderboard v2	38%	#14 of 21	Official leaderboard
LLM Chess	-289	#101 of 124	Official leaderboard
Terminal-Bench	17.1%	#16 of 18	Epoch AI
Terminal-Bench 2.0	17.1%	#41 of 43	Official leaderboard

Professional work

Benchmark	Score	Rank	Source
ProfBench	49.2%	#31 of 72	Official leaderboard
MedScribe	83%	#45 of 98	Official leaderboard
MedCode	40.4%	#62 of 96	Official leaderboard
PRBench	41%	#29 of 44	Official leaderboard
Vending-Bench 2	$549	#54 of 65	Epoch AI

Vision, audio and robotics

Benchmark	Score	Rank	Source
Math-VR	60.5%	#4 of 26	Official leaderboard
MMMU	79.7%	#21 of 97	Model card
ActivityNet-QA	65.1%	#2 of 8	Model card
MADQA	71.8%	#5 of 16	Official leaderboard
Perception Test MCVQA	75.1%	#2 of 6	Model card
SpatialViz-Bench	36.9%	#3 of 8	Epoch AI
Vibe-Eval	65.4%	#4 of 9	Model card
Video-MME	81.5%	#21 of 42	Model card
Audio MC	40%	#8 of 16	Official leaderboard
VATEX	65.2	#2 of 4	Model card
YouCook2	177.6	#2 of 4	Model card
Vision Arena	1,215	#69 of 136	Official leaderboard
BetterChartQA	67.3%	#2 of 3	Model card
Video-MMMU	79.2%	#23 of 30	Model card
LVBench	62.7%	#20 of 25	Model card
CoVoST2	36.1	#9 of 11	Model card
ZeroBench	2.00%	#24 of 25	Model card
MINERVA	60.7%	#6 of 6	Model card
EgoTempo	36.7%	#5 of 5	Model card
QVHighlights	52.4%	#5 of 5	Model card
VATEX ZH	43.9	#4 of 4	Model card
1H-VideoQA	67.5%	#2 of 2	Model card
Neptune	84.3%	#2 of 2	Model card

Safety and security

Benchmark	Score	Rank	Source
Hack the Box challenge set (Google DeepMind)	15.4%	#1 of 4	Model card
InterCode-CTF	98.7%	#2 of 4	Model card
Divergence attack (extractable memorization)	59%	#1 of 2	Model card
Indirect prompt injection, Actor Critic attack	40.8%	#1 of 2	Model card
Indirect prompt injection, Beam Search attack	4.20%	#1 of 2	Model card
Human Pathogen Capabilities Test v2 (HPCT v2)	43.5%	#52 of 71	Official leaderboard
In-house CTF suite (Phuong et al. 2024)	76.9%	#3 of 4	Model card
WMDP Chemistry	75%	#4 of 5	Model card
LAB-Bench CloningScenarios	44%	#9 of 11	Model card
WMDP Biology	84%	#6 of 7	Model card
LAB-Bench ProtocolQA	64%	#15 of 15	Model card
LAB-Bench SeqQA	58%	#12 of 12	Model card
Cyber key skills benchmark v1	8.30%	#4 of 4	Model card
SecureBio VMQA, single choice	48%	#4 of 4	Model card
Dangerous Content policy violations (from ART)	26.9%	#2 of 2	Model card
Helpfulness violations (from ART)	6.60%	#2 of 2	Model card
Indirect prompt injection, TAP attack	53.6%	#2 of 2	Model card

Other

Benchmark	Score	Rank	Source
SAGE (Vals)	44.8%	#44 of 81	Official leaderboard
LMCA	27.5%	#91 of 139	Epoch AI
Conceptual Reasoning Index	37.7	#102 of 149	Official leaderboard
CAIS Vision Capabilities Index	41.1	#42 of 53	Official leaderboard
ACCoRD	45.4	#120 of 150	Official leaderboard
CAIS Text Capabilities Index	13.2	#58 of 66	Official leaderboard
CAIS Risk Index	70.0	#63 of 65	Official leaderboard

Google DeepMind's models over time

Google DeepMind's models in color, against every other model on the Epoch Capabilities Index, by release date.

Its launch post

The benchmarks the lab chose to headline, with the scores it gave.

Gemini 2.5 Flash Model Card, GA Thinking column (Google DeepMind) 17 Jun 2025

Humanity's Last Exam (no tools), GA Thinking 11
GPQA diamond single attempt 82.8
AIME 2025 single attempt 72
LiveCodeBench v5 single attempt 63.9
Aider Polyglot whole 61.9 / diff 56.7 61.9
SWE-Bench Verified 60.4
SimpleQA 26.9
FACTS Grounding 85.3
MMMU single attempt 79.7
Vibe-Eval (Reka) 65.4
MRCR v2 128k (average), earlier non-8-needle version (1M pointwise 32) 74
Global MMLU (Lite) 88.4

Its system card

Every result in the longer documents the lab published with it. 45 of them aren't in the launch post.

Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities 7 Jul 2025