Head-to-head comparison

Gemini 3.8 Flash vs. Gemini 3.5 Flash-Lite

Compare published benchmark results from matching versions and cohorts, API pricing, and technical specifications. Any documented test setup differences remain visible.

shared result series: 207
distinct benchmarks: 25
latest retrieval: 2026-10-15

First modelSecond model

Specifications

General specifications

Release date, availability, context window, architecture, and published parameter counts at a glance. Undisclosed details remain marked as unavailable.

Specifications for Gemini 3.8 Flash and Gemini 3.5 Flash-Lite
Attribute	Gemini 3.8 Flash	Gemini 3.5 Flash-Lite
Provider	Google	Google
Released	Sep 2, 2026	Jul 21, 2026
Availability	Active	Active
Model type	Proprietary	Proprietary
Parameters	Not published	Not published
Architecture	Not published	Not published
Context window	1,048,576 tokens	1,048,576 tokens
Knowledge cutoff	March 2026	Not published
Technical sources	Model, Knowledge cutoff	Model

Pricing

API pricing compared

The table compares official standard API rates per 1M tokens, including published cache pricing. Batch, fast, priority, regional surcharges, tool calls, and cloud platform pricing are excluded.

API pricing for Gemini 3.8 Flash and Gemini 3.5 Flash-Lite
Price type	Gemini 3.8 Flash	Gemini 3.5 Flash-Lite
API input	$0.75	$0.3
API output	$3.75	$2.5
Cache read	$0.075	$0.03
Cache write (5 min.)	Not listed	Not listed
Cache write (1 hr.)	Not listed	Not listed
Price verified	09/22/2026 Source	09/22/2026 Source

Gemini 3.8 FlashGemini 3.5 Flash-Lite

API inputUSD per 1M tokens, base tier

Gemini 3.8 Flash$0.75

Gemini 3.5 Flash-Lite$0.3

API outputUSD per 1M tokens, base tier

Gemini 3.8 Flash$3.75

Gemini 3.5 Flash-Lite$2.5

Cache readUSD per 1M tokens, base tier

Gemini 3.8 Flash$0.075

Gemini 3.5 Flash-Lite$0.03

Key to the marks:better valuebetter, but not conclusivetie

Pricing by provider

Direct developer pricing and provider endpoint tariffs routed through OpenRouter are listed separately. Prices are in USD per 1M tokens. Context-dependent direct tiers plus regional, Flex, Priority, and other OpenRouter tariffs remain individually identifiable.

Provider pricing for Gemini 3.8 Flash and Gemini 3.5 Flash-Lite
Model	Provider	Purchase route and tariff	Input per 1M tokens	Output per 1M tokens	Source
Gemini 3.8 Flash	Google	Direct from developerStandard API	$0.75	$3.75	Source09/22/2026
	Google	Via OpenRoutergoogle-vertex/us	$0.825	$4.125	Source10/17/2026
	Google AI Studio	Via OpenRoutergoogle-ai-studio	$0.75	$3.75	Source10/17/2026
	Google AI Studio	Via OpenRoutergoogle-ai-studio/flex	Lowest listed price: $0.375	Lowest listed price: $1.875	Source10/17/2026
	Google AI Studio	Via OpenRoutergoogle-ai-studio/priority	$1.35	$6.75	Source10/17/2026
Gemini 3.5 Flash-Lite	Google	Direct from developerStandard API	$0.3	$2.5	Source09/22/2026
	Google	Via OpenRoutergoogle-vertex/eu	$0.33	$2.75	Source10/17/2026
	Google	Via OpenRoutergoogle-vertex/global	$0.3	$2.5	Source10/17/2026
	Google	Via OpenRoutergoogle-vertex/global/priority	$0.54	$4.5	Source10/17/2026
	Google	Via OpenRoutergoogle-vertex/us	$0.33	$2.75	Source10/17/2026
	Google AI Studio	Via OpenRoutergoogle-ai-studio	$0.3	$2.5	Source10/17/2026
	Google AI Studio	Via OpenRoutergoogle-ai-studio/flex	Lowest listed price: $0.15	Lowest listed price: $1.25	Source10/17/2026
	Google AI Studio	Via OpenRoutergoogle-ai-studio/priority	$0.54	$4.5	Source10/17/2026

Performance

Shared benchmarks

Only published results with the same benchmark version, task, metric, and comparison cohort are paired. Different reasoning levels, agents, harnesses, or output limits appear directly in the table.

Capability profile from matched benchmarks

Each axis averages directly matched benchmark families. A value of 100 means the stronger value within this pair, not a universal quality score.

Gemini 3.8 Flash

Gemini 3.5 Flash-Lite

Coding: 100 / 53.7 (5 benchmark families)

Reasoning: 100 / 88.8 (1 benchmark family)

Knowledge: 100 / 95.1 (1 benchmark family)

Finance: 99.1 / 74.6 (5 benchmark families)

Legal: 100 / 75 (3 benchmark families)

Agentic tasks: 100 / 60.4 (1 benchmark family)

Multimodal: 100 / 93.8 (1 benchmark family)

Productivity: 100 / 79.4 (1 benchmark family)

165 matched rows from 20 benchmark families were considered. The radar shows 8 of 10 comparable categories.

Gemini 3.8 FlashGemini 3.5 Flash-Lite

Code MigrationOverall. Higher is better. No documented setup difference

Gemini 3.8 Flash36.546 %

Gemini 3.5 Flash-Lite6.148 %

CyberBench v1.1Overall. Higher is better. No documented setup difference

Gemini 3.8 Flash43.75 %

Gemini 3.5 Flash-Lite59.107 %

EMBOverall. Higher is better. No documented setup difference

Gemini 3.8 Flash72.196 %

Gemini 3.5 Flash-Lite43.231 %

Finance Agent (v2)Overall. Higher is better. No documented setup difference

Gemini 3.8 Flash61.435 %

Gemini 3.5 Flash-Lite47.442 %

GPQA DiamondOverall. Higher is better. No documented setup difference

Gemini 3.8 Flash94.444 %

Gemini 3.5 Flash-Lite83.838 %

Harvey's Legal Agent BenchmarkOverall · Task fully resolved. Higher is better. No documented setup difference

Gemini 3.8 Flash10 %

Gemini 3.5 Flash-Lite0 %

Legal Research BenchOverall · All-pass. Higher is better. No documented setup difference

Gemini 3.8 Flash38.942 %

Gemini 3.5 Flash-Lite13.942 %

LegalBenchOverall. Higher is better. No documented setup difference

Gemini 3.8 Flash86.993 %

Gemini 3.5 Flash-Lite84.073 %

LiveCodeBench v6Overall. Higher is better. No documented setup difference

Gemini 3.8 Flash89.484 %

Gemini 3.5 Flash-Lite79.014 %

LMArena Code Arena WebDevOverall. Higher is better. Test setup not documented as identical

Gemini 3.8 Flash1,584.77

Gemini 3.5 Flash-Lite1,440.401

LMArena Text, Style ControlOverall. Higher is better. Test setup not documented as identical

Gemini 3.8 Flash1,496.319

Gemini 3.5 Flash-Lite1,454.971

LMArena Vision, Style ControlOverall. Higher is better. Test setup not documented as identical

Gemini 3.8 Flash1,289.129

Gemini 3.5 Flash-Lite1,265.999

Key to the marks:better valuebetter, but not conclusivetie

Benchmark scores for Gemini 3.8 Flash and Gemini 3.5 Flash-Lite
Benchmark and task	Gemini 3.8 Flash	Gemini 3.5 Flash-Lite	Source
Code MigrationOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner36.546 %	6.148 %	Vals AI Source details Data as of: 10/12/2026
CyberBench v1.1Overall Test details No documented difference in the test setupVersion: 1.1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	43.75 %	Winner59.107 %	Vals AI Source details Data as of: 10/12/2026
EMBOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner72.196 %	43.231 %	Vals AI Source details Data as of: 10/12/2026
Finance Agent (v2)Overall Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner61.435 %	47.442 %	Vals AI Source details Data as of: 10/12/2026
GPQA DiamondOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner94.444 %	83.838 %	Vals AI Source details Data as of: 09/01/2026
Harvey's Legal Agent BenchmarkOverall · Task fully resolved Test details No documented difference in the test setupVersion: 1Metric: task resolution rateScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner10 %	0 %	Vals AI Source details Data as of: 10/12/2026
Legal Research BenchOverall · All-pass Test details No documented difference in the test setupVersion: 1Metric: all-pass rateScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner38.942 %	13.942 %	Vals AI Source details Data as of: 10/12/2026
LegalBenchOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner86.993 %	84.073 %	Vals AI Source details Data as of: 10/13/2026
LiveCodeBench v6Overall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner89.484 %	79.014 %	Vals AI Source details Data as of: 09/01/2026
LMArena Code Arena WebDevOverall Test details Test setup not documented as identicalNo direct winnerVersion: webdevMetric: WebDev Arena Bradley-Terry ratingScoring: Higher is betterGemini 3.8 Flash: Snapshot: gemini-3.8-flash-high, Reasoning: high, Harness: Arena Code Arena WebDev Gemini 3.5 Flash-Lite: Snapshot: gemini-3.5-flash-lite, Harness: Arena Code Arena WebDevGemini 3.8 Flash: Arena rating from 10,077 published pairwise votes.Gemini 3.5 Flash-Lite: Arena rating from 242 published pairwise votes.	Higher measured value1,584.77 95% confidence interval: ±7.59The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,440.401 95% confidence interval: ±40.572The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/13/2026
LMArena Text, Style ControlOverall Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGemini 3.8 Flash: Snapshot: gemini-3.8-flash-high, Reasoning: high, Harness: Arena Text, style control Gemini 3.5 Flash-Lite: Snapshot: gemini-3.5-flash-lite, Harness: Arena Text, style controlGemini 3.8 Flash: Arena rating from 35,670 published pairwise votes.Gemini 3.5 Flash-Lite: Arena rating from 35,845 published pairwise votes.	Higher measured value1,496.319 95% confidence interval: ±4.654The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,454.971 95% confidence interval: ±4.474The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/13/2026
LMArena Vision, Style ControlOverall Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGemini 3.8 Flash: Snapshot: gemini-3.8-flash-high, Reasoning: high, Harness: Arena Vision, style control Gemini 3.5 Flash-Lite: Snapshot: gemini-3.5-flash-lite, Harness: Arena Vision, style controlGemini 3.8 Flash: Arena rating from 3,792 published pairwise votes.Gemini 3.5 Flash-Lite: Arena rating from 8,381 published pairwise votes.	Higher measured value1,289.129 95% confidence interval: ±10.62The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,265.999 95% confidence interval: ±8.109The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/09/2026
MedCodeOverall Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterGemini 3.8 Flash: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens Gemini 3.5 Flash-Lite: Reasoning: high, Temperature: 1, Output limit: 30,000 tokens	Higher measured value48.135 %	43.489 %	Vals AI Source details Data as of: 10/13/2026
MedScribeOverall Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterGemini 3.8 Flash: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens Gemini 3.5 Flash-Lite: Reasoning: high, Temperature: 1, Output limit: 30,000 tokens	Higher measured value84.496 %	70.886 %	Vals AI Source details Data as of: 10/13/2026
MMLU-ProOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner90.221 %	85.836 %	Vals AI Source details Data as of: 09/01/2026
MMMU-ProOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner89.075 %	83.584 %	Vals AI Source details Data as of: 09/01/2026
MortgageTaxOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	65.342 %	Winner68.68 %	Vals AI Source details Data as of: 09/01/2026
ProgramBenchOverall Test details No documented difference in the test setupVersion: 1Metric: fully resolved rateScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner1 %	0 %	Vals AI Source details Data as of: 10/12/2026
Public Benefits Bench v1.1Overall Test details No documented difference in the test setupVersion: 1.1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner65.291 %	51.827 %	Vals AI Source details Data as of: 10/12/2026
SAGEOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 30,000 tokens	35.064 %	Winner47.303 %	Vals AI Source details Data as of: 10/08/2026
SWE-bench VerifiedOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner80 %	75 %	Vals AI Source details Data as of: 09/01/2026
Tax Agent BenchOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner32.124 %	7.254 %	Vals AI Source details Data as of: 10/12/2026
TaxEval v2Overall Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner74.448 %	72.608 %	Vals AI Source details Data as of: 09/01/2026
Terminal-Bench 2.1Overall Test details No documented difference in the test setupVersion: 2.1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner81.273 %	50.187 %	Vals AI Source details Data as of: 09/28/2026
Vibe Code Bench v1.1Overall Test details No documented difference in the test setupVersion: 1.1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Harness: OpenHands, Temperature: 1, Output limit: 65,536 tokens	Winner78.651 %	37.161 %	Vals AI Source details Data as of: 10/12/2026
Code MigrationCLI Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner25.871 %	5.875 %	Vals AI Source details Data as of: 10/12/2026
CyberBench v1.1Patch Test details No documented difference in the test setupVersion: 1.1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner87.5 %	73.214 %	Vals AI Source details Data as of: 10/12/2026
EMBComps Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner76.435 %	46.417 %	Vals AI Source details Data as of: 10/12/2026
Finance Agent (v2)Adjustments Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner54.828 %	41.741 %	Vals AI Source details Data as of: 10/12/2026
GPQA DiamondFew-Shot CoT Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner94.949 %	81.818 %	Vals AI Source details Data as of: 09/01/2026
Harvey's Legal Agent BenchmarkAntitrust Competition · Criteria passed Test details No documented difference in the test setupVersion: 1Metric: criteria pass rateScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner87.681 %	76.449 %	Vals AI Source details Data as of: 10/12/2026
Legal Research BenchAdministrative / Regulatory · All-pass Test details No documented difference in the test setupVersion: 1Metric: all-pass rateScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner46.97 %	19.697 %	Vals AI Source details Data as of: 10/12/2026
LegalBenchConclusion Tasks Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner90.691 %	88.404 %	Vals AI Source details Data as of: 10/13/2026
LiveCodeBench v6Easy Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner99.068 %	98.758 %	Vals AI Source details Data as of: 09/01/2026
LMArena Code Arena WebDevImage to webdev Test details Test setup not documented as identicalNo direct winnerVersion: webdevMetric: WebDev Arena Bradley-Terry ratingScoring: Higher is betterGemini 3.8 Flash: Snapshot: gemini-3.8-flash-high, Reasoning: high, Harness: Arena Code Arena WebDev Gemini 3.5 Flash-Lite: Snapshot: gemini-3.5-flash-lite, Harness: Arena Code Arena WebDevGemini 3.8 Flash: Arena rating from 2,015 published pairwise votes.Gemini 3.5 Flash-Lite: Arena rating from 1,600 published pairwise votes.	Higher measured value1,593.067 95% confidence interval: ±13.862The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,447.43 95% confidence interval: ±15.334The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/09/2026
LMArena Text, Style ControlBusiness, management and finance Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGemini 3.8 Flash: Snapshot: gemini-3.8-flash-high, Reasoning: high, Harness: Arena Text, style control Gemini 3.5 Flash-Lite: Snapshot: gemini-3.5-flash-lite, Harness: Arena Text, style controlGemini 3.8 Flash: Arena rating from 6,520 published pairwise votes.Gemini 3.5 Flash-Lite: Arena rating from 6,603 published pairwise votes.	Higher measured value1,479.887 95% confidence interval: ±8.136The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,452.709 95% confidence interval: ±7.982The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/13/2026
LMArena Vision, Style ControlChinese Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGemini 3.8 Flash: Snapshot: gemini-3.8-flash-high, Reasoning: high, Harness: Arena Vision, style control Gemini 3.5 Flash-Lite: Snapshot: gemini-3.5-flash-lite, Harness: Arena Vision, style controlGemini 3.8 Flash: Arena rating from 182 published pairwise votes.Gemini 3.5 Flash-Lite: Arena rating from 471 published pairwise votes.	Higher measured value1,379.729 95% confidence interval: ±43.521The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,301.978 95% confidence interval: ±28.157The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/09/2026
MMLU-ProBiology Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner94.979 %	91.911 %	Vals AI Source details Data as of: 09/01/2026
MortgageTaxNumerical Extraction Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	70.588 %	Winner75.835 %	Vals AI Source details Data as of: 09/01/2026
ProgramBenchAlmost Resolved Test details No documented difference in the test setupVersion: 1Metric: almost-resolved rateScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner14 %	0 %	Vals AI Source details Data as of: 10/12/2026
SAGEAutomata Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 30,000 tokens	Winner33.334 %	28.968 %	Vals AI Source details Data as of: 10/08/2026
SWE-bench Verified<15 min fix Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner89.691 %	85.567 %	Vals AI Source details Data as of: 09/01/2026
Tax Agent BenchControversy & Precedence Analysis Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner6.667 %	3.333 %	Vals AI Source details Data as of: 10/12/2026
TaxEval v2Correctness Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner62.469 %	61.733 %	Vals AI Source details Data as of: 09/01/2026
Terminal-Bench 2.1Easy Test details No documented difference in the test setupVersion: 2.1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner100 %	66.667 %	Vals AI Source details Data as of: 09/28/2026
Code MigrationCOBOL Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner68.571 %	6.964 %	Vals AI Source details Data as of: 10/12/2026
CyberBench v1.1PoC Test details No documented difference in the test setupVersion: 1.1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	0 %	Winner45 %	Vals AI Source details Data as of: 10/12/2026
EMBDataroom Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner86.218 %	57.687 %	Vals AI Source details Data as of: 10/12/2026
Finance Agent (v2)All-Pass Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner49.694 %	35.265 %	Vals AI Source details Data as of: 10/12/2026
GPQA DiamondZero-Shot CoT Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner93.939 %	85.859 %	Vals AI Source details Data as of: 09/01/2026
Harvey's Legal Agent BenchmarkAntitrust Competition · Task fully resolved Test details No documented difference in the test setupVersion: 1Metric: task resolution rateScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Tie0 %	Tie0 %	Vals AI Source details Data as of: 10/12/2026
Legal Research BenchAdministrative / Regulatory · Weighted partial credit Test details No documented difference in the test setupVersion: 1Metric: weighted partial-credit scoreScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner87.613 %	71.96 %	Vals AI Source details Data as of: 10/12/2026
LegalBenchInterpretation Tasks Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner84.589 %	80.332 %	Vals AI Source details Data as of: 10/13/2026
LiveCodeBench v6Hard Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner78 %	53.429 %	Vals AI Source details Data as of: 09/01/2026
LMArena Code Arena WebDevWebdev Test details Test setup not documented as identicalNo direct winnerVersion: webdevMetric: WebDev Arena Bradley-Terry ratingScoring: Higher is betterGemini 3.8 Flash: Snapshot: gemini-3.8-flash-high, Reasoning: high, Harness: Arena Code Arena WebDev Gemini 3.5 Flash-Lite: Snapshot: gemini-3.5-flash-lite, Harness: Arena Code Arena WebDevGemini 3.8 Flash: Arena rating from 10,077 published pairwise votes.Gemini 3.5 Flash-Lite: Arena rating from 242 published pairwise votes.	Higher measured value1,584.77 95% confidence interval: ±7.59The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,440.401 95% confidence interval: ±40.572The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/13/2026
LMArena Text, Style ControlChinese Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGemini 3.8 Flash: Snapshot: gemini-3.8-flash-high, Reasoning: high, Harness: Arena Text, style control Gemini 3.5 Flash-Lite: Snapshot: gemini-3.5-flash-lite, Harness: Arena Text, style controlGemini 3.8 Flash: Arena rating from 2,903 published pairwise votes.Gemini 3.5 Flash-Lite: Arena rating from 2,746 published pairwise votes.	Higher measured value1,532.833 95% confidence interval: ±11.757The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,483.866 95% confidence interval: ±12.01The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/13/2026
LMArena Vision, Style ControlCreative writing vision Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGemini 3.8 Flash: Snapshot: gemini-3.8-flash-high, Reasoning: high, Harness: Arena Vision, style control Gemini 3.5 Flash-Lite: Snapshot: gemini-3.5-flash-lite, Harness: Arena Vision, style controlGemini 3.8 Flash: Arena rating from 363 published pairwise votes.Gemini 3.5 Flash-Lite: Arena rating from 787 published pairwise votes.	Higher measured value1,263.378 95% confidence interval: ±32.508The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,256.499 95% confidence interval: ±22.214The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/09/2026
MMLU-ProBusiness Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner93.79 %	90.621 %	Vals AI Source details Data as of: 09/01/2026
MortgageTaxSemantic Extraction Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	60.095 %	Winner61.526 %	Vals AI Source details Data as of: 09/01/2026
ProgramBenchFully Resolved Test details No documented difference in the test setupVersion: 1Metric: fully resolved rateScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner1 %	0 %	Vals AI Source details Data as of: 10/12/2026
SAGECalculus Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 30,000 tokens	Winner79.63 %	60.494 %	Vals AI Source details Data as of: 10/08/2026
SWE-bench Verified>4 hours Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Tie33.333 %	Tie33.333 %	Vals AI Source details Data as of: 09/01/2026
Tax Agent BenchControversy & Precedence Analysis Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner59.001 %	34.598 %	Vals AI Source details Data as of: 10/12/2026
TaxEval v2Stepwise Reasoning Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterSettings: Reasoning: high, Temperature: 1, Output limit: 65,536 tokens	Winner86.427 %	83.483 %	Vals AI Source details Data as of: 09/01/2026
Key to the marks:better valuebetter, but not conclusivetie

Our model tests

Our tests

SVG short test

One SVG prompt and fixed output settings. Captures are shown only with 0 reported reasoning tokens or documented lack of reasoning controls. Identical test conditions do not make the models' internal compute identical.

Shared test round: 4ae5d8eeb9a9

Gemini 3.8 Flash

Unavailable in this round

Gemini 3.5 Flash-Lite

Unavailable in this round

Earlier tests

Pelican on a bicycle

A community classic for free-form SVG drawing.

Gemini 3.8 Flash

The response reached the token limit. No complete recording is available.

Reasoning (requested): high

Current attempt

Run date: Oct 5, 2026
API provider: openrouter
Requested API model: google/gemini-3.8-flash
API model reported in response: google/gemini-3.8-flash
Reasoning setting: high
Pinned endpoint: Google AI Studio | google/gemini-3.8-flash-20260902
Temperature: Provider default
Token limit including reasoning: 8,192 tokens
Test protocol: community-visual-tests-v1
Technical checks: Failed
Response time: 40.2 s

Not yet visually reviewed. A successful recording does not confirm correct physics or full compliance with the prompt.

Earlier attempts (3)

Oct 2, 2026API request failedReasoning setting: high
Oct 1, 2026API request failedReasoning setting: high
Sep 7, 2026API request failedReasoning setting: high

Gemini 3.5 Flash-Lite

Reasoning (requested): high

Current attempt

Run date: Oct 2, 2026
API provider: openrouter
Requested API model: google/gemini-3.5-flash-lite
API model reported in response: google/gemini-3.5-flash-lite
Reasoning setting: high
Pinned endpoint: Google AI Studio | google/gemini-3.5-flash-lite-20260721
Temperature: Provider default
Token limit including reasoning: 8,192 tokens
Test protocol: community-visual-tests-v1
Technical checks: Passed, not a quality rating
Response time: 14.3 s

Visually reviewed. This individual test does not establish overall model quality.

Earlier attempts (2)

Oct 1, 2026API request failedReasoning setting: high
Sep 7, 2026API request failedReasoning setting: high

Prompt and test conditions

Generate an SVG of a pelican riding a bicycle

Token limit including reasoning: 8,192 tokens

Task origin (Simon Willison)

How to read this comparison

Winning one benchmark is not an overall verdict

Your workload, budget, and required context length matter most. A coding benchmark says little about visual understanding or agent performance.

The standard price table uses direct developer rates. The provider comparison labels OpenRouter endpoints, regions, and special tariffs separately.

Every result links to its measurement source and retrieval date. Labels distinguish vendor reports, official benchmark leaderboards, and independent evaluations. Disputed or archived results are excluded.

More matchups

Compare Gemini 3.8 Flash and Gemini 3.5 Flash-Lite with other leading models using the same data and benchmark logic.

View the complete LLM comparison

Gemini 3.8 Flash vs. Gemini 3.5 Flash-Lite

Change model selection

General specifications

API pricing compared

Shared benchmarks

Capability profile from matched benchmarks

Our tests

SVG short test

Gemini 3.8 Flash

Gemini 3.5 Flash-Lite

Pelican on a bicycle

Winning one benchmark is not an overall verdict

Related LLM comparisons