GoogleReasoning modelAuto-detected

Gemini 3.7 Flash

Name: Gemini 3.7 Flash benchmark results
Creator: BenchLeader

Best configuration ranks #34 of 610 on the BenchLeader Index at 64.4 ±3.9 (medium reasoning effort). Last measured 13 Aug 2026. Released 13 Aug 2026.

Blended price: $1.50/M; $0.750 in · $3.75 out
Output speed: 282 tok/s
First answer: 5.27 s; first token 1.55 s
Context: 1M

How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.

Overall index64
Reasoning64
Coding68
Knowledge75
Multimodal69
Long context68
Composite79

Reasoning-effort configurations

The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Default” means the publisher did not say which setting was used.

Effort	Index	Rank	Speed	First answer	Chat reply cost	Categories
low	62.5	#53	303 tok/s	0.62 s	$0.0014	Coding 63 · Composite 76 · Knowledge 75 · Long context 66 · Multimodal 70 · Reasoning 59
mediumbest	64.4	#34	282 tok/s	5.27 s	$0.0014	Coding 68 · Composite 79 · Knowledge 75 · Long context 68 · Multimodal 69 · Reasoning 64
high	60.8	#75	297 tok/s	9.42 s	$0.0014	Agents & tools 40 · Coding 66 · Composite 63 · Human preference 69 · Knowledge 67 · Maths 63 · Reasoning 69
default	61.7	#63	297 tok/s	9.42 s	$0.0014	Agents & tools 52 · Coding 60 · Composite 79 · Knowledge 77 · Long context 67 · Maths 62 · Multimodal 70

Benchmark results

One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.

Reasoning

Benchmark	low	medium	high	default	Source
GPQA Diamond	–	–	94.8%#3	–	Epoch AI Benchmarking Hub
LMArena Hard Prompts	–	–	1508#14	–	LMArena
LiveBench Reasoningnot in index	–	–	87.8%#18	–	LiveBench
GPQA Diamond (AA)not in index	90.1%#64	92.1%#39	–	94.5%#6	Artificial Analysis
Humanity's Last Exam (AA)not in index	35.1%#80	39.0%#61	–	47.9%#19	Artificial Analysis
GPQA Diamond (Vals)not in index	–	–	93.9%#5	–	Vals AI
ARC-AGI-1	85.2%#75	91.2%#45	95.5%#20	–	ARC Prize
ARC-AGI-2	52.9%#69	63.8%#50	84.6%#19	–	ARC Prize

Coding

Benchmark	low	medium	high	default	Source
SciCode	53.6%#37	57.9%#8	56.8%#11	–	SciCode
FrontierCode	–	–	–	43.6%#9	Cognition
LMArena Coding	–	–	1522#23	–	LMArena
LMArena WebDev	–	–	1587#17	–	LMArena
LiveBench Codingnot in index	–	–	78.9%#19	–	LiveBench
SciCode (AA)not in index	55.7%#28	59.8%#4	–	57.2%#15	Artificial Analysis
LiveCodeBench	–	–	88.7%#5	–	Vals AI
SWE-bench (Vals)not in index	–	–	80.8%#23	–	Vals AI

Agents & tools

Benchmark	low	medium	high	default	Source
Terminal-Bench	–	–	11.2%#64	–	Terminal-Bench
Remote Labor Index	–	–	–	5.0%#4	Scale AI / CAIS
LiveBench Agentic Codingnot in index	–	–	58.3%#15	–	LiveBench
Terminal-Bench 2.1 (Vals)	–	–	77.5%#11	–	Vals AI

Maths

Benchmark	low	medium	high	default	Source
FrontierMath Tiers 1–3	–	–	71.6%#16	–	Epoch AI Benchmarking Hub
FrontierMath Tier 4	–	–	36.6%#23	–	Epoch AI Benchmarking Hub
OTIS Mock AIME	–	–	97.2%#23	–	Epoch AI Benchmarking Hub
ProofBench	–	–	–	58.0%#11	Vals AI
LiveBench Mathematicsnot in index	–	–	93.5%#12	–	LiveBench

Knowledge

Benchmark	low	medium	high	default	Source
SimpleQA Verified	–	–	69.2%#7	–	Epoch AI Benchmarking Hub
LiveBench Data Analysisnot in index	–	–	68.0%#42	–	LiveBench
AA-Omniscience	22.1#34	23.7#32	–	26.5#28	Artificial Analysis
MMLU-Pro	–	–	90.1%#6	–	Vals AI
LegalBench	–	–	87.3%#4	–	Vals AI
TaxEval	–	–	74.7%#30	–	Vals AI

Instruction following

Benchmark	low	medium	high	default	Source	Trend
LiveBench Languagenot in index	–	–	85.5%#9	–	LiveBench

Human preference

Benchmark	low	medium	high	default	Source	Trend
LMArena Text	–	–	1491#11	–	LMArena

Multimodal

Benchmark	low	medium	high	default	Source	Trend
MMMU-Pro	84.9%#7	84.7%#8	–	85.5%#5	Artificial Analysis

Long context

Benchmark	low	medium	high	default	Source	Trend
AA-LCR	78.7%#75	83.0%#12	–	81.7%#30	Artificial Analysis

Composite

Benchmark	low	medium	high	default	Source
Epoch Capabilities Indexnot in index	–	–	–	157.4#12	Epoch AI Benchmarking Hub
LiveBench	–	–	78.8%#9	–	LiveBench
AA Intelligence Index	37.0#49	39.6#38	–	39.4#40	Artificial Analysis
Vals Indexnot in index	–	–	59.3#13	–	Vals AI

Where to run it

Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.

Provider	Speed	First token	Input $/M	Output $/M	Context	Quantisation
Google AI Studio Priority	190 tok/s	1.33 s	$1.35	$6.75	1.0M	–
Google AI Studio Flex	175 tok/s	1.12 s	$0.375	$1.88	1.0M	–
Google AI Studio	162 tok/s	1.34 s	$0.750	$3.75	1.0M	–
Google Vertex	83 tok/s	2.27 s	$0.750	$3.75	1.0M	–
Google Vertex Priority	79 tok/s	1.75 s	$1.35	$6.75	1.0M	–
Google Vertex Flex	27 tok/s	13 s	$0.375	$1.88	1.0M	–

Price history

Listed price per 1M tokens over time, as recorded by OpenRouter for the provider with the longest history.

input outputnow $0.750 in · $3.75 out

What a task costs

Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.075 per 1M. Reasoning tokens are not modelled.

Workload	Tokens in / out	Cost	With caching	Time
Chat reply	400 / 300	$0.0014	$0.0012	6.3 s
Summarise a 30-page report	12,000 / 600	$0.011	$0.0052	7.4 s
Code edit	6,000 / 1,500	$0.010	$0.0071	10.6 s
Agentic coding session	60,000 / 4,000	$0.060	$0.030	19.4 s
Structured extraction	2,000 / 200	$0.0023	$0.0012	6.0 s