BenchLeader

Gemini 3.7 Flash vs Step 5 Preview

Full profiles: Gemini 3.7 Flash (Google) and Step 5 Preview (StepFun): every benchmark, price and rank.

Updated · Data from 10 sources

As of 16 Nov 2026, Gemini 3.7 Flash is ahead of Step 5 Preview on quality (62.1 vs 56.0 on the BenchLeader Index), and they cost about the same at $1.50 vs $1.43 per 1M tokens blended. Choose Gemini 3.7 Flash for reasoning (68 vs 52), coding (62 vs 50), and knowledge (66 vs 54) and fast first answers (2.2 s vs 23.5 s); choose Step 5 Preview for long context (69 vs 66), where it also streams ~1.5× faster.

Key facts (as of 16 Nov 2026)

  • Gemini 3.7 Flash scores 62.1 on the BenchLeader Index and Step 5 Preview scores 56.0 (BenchLeader, 16 Nov 2026).
  • Gemini 3.7 Flash costs $1.50 per 1M tokens blended (list price, 16 Nov 2026).
  • Step 5 Preview costs $1.43 per 1M tokens blended (list price, 16 Nov 2026).
  • Step 5 Preview streams ~1.5× faster than Gemini 3.7 Flash (Artificial Analysis, 7-day median, 16 Nov 2026).
  • Of the 35 benchmarks both models have, Gemini 3.7 Flash leads 20 (BenchLeader, 16 Nov 2026).

Which to pick for…

The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.

Use casePickWhyConfidence
CodingGemini 3.7 Flash62 vs 50 category score; Arena Coding 1520 vs 1505High
Agents & tool useEither49 vs 49 category score; Arena Agent -2.2 vs -3.7Either
MathsGemini 3.7 Flash63 vs 60 category score; Arena Maths 1502 vs 1470Medium
ReasoningGemini 3.7 Flash68 vs 52 category score; Arena Hard Prompts 1510 vs 1477High
KnowledgeGemini 3.7 Flash66 vs 54 category score; Excel Modeling Benchmark 71.3% vs 57.1%High
Long documentsStep 5 Preview69 vs 66 category score; AA-LCR v1.1 88.3% vs 81.7%Medium
Images and other mediaGemini 3.7 Flash67 vs 59 category score; Arena Vision 1313 vs 1263High
Chat quality (human preference)Gemini 3.7 Flash68 vs 63 on Arena votesMedium
Lowest costEither$1.50 vs $1.43 per 1M blended; agentic coding session $0.060 vs $0.071–
Fastest first answerGemini 3.7 Flash2.2 s (Gemini 3.7 Flash (high)) vs 23.5 s (Step 5 Preview) to the first answer token–
High volume on a budgetNeither flagshipA smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs–

Key differences

  • Gemini 3.7 Flash (high) leads on quality: 62.1 vs 56.0.
  • Gemini 3.7 Flash (high) is stronger in coding, human preference, knowledge, maths, multimodal, reasoning.
  • Step 5 Preview is stronger in agents & tools, long context.
  • They cost about the same ($1.43 per 1M blended).
  • Step 5 Preview streams 1.3× faster (124 vs 95 tokens per second).

At a glance

SpecGemini 3.7 FlashStep 5 Preview
DeveloperGoogleStepFun
Released13 Aug 202618 Sep 2026
Context window (tokens)1,048,5761,000,000
Compared atGemini 3.7 Flash (high)Step 5 Preview
Open weightsNoYes
Price in / out / cached input per 1M$0.750 / $3.75 / $0.075$1.00 / $2.70 / $0.105

Category by category

Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.

CategoryGemini 3.7 Flash (high)Step 5 PreviewLeader
Reasoning67.851.7Gemini 3.7 Flash
Coding62.349.9Gemini 3.7 Flash
Agents & tools48.949.1level
Maths62.660.1Gemini 3.7 Flash
Knowledge65.953.7Gemini 3.7 Flash
Human preference68.163.1Gemini 3.7 Flash
Multimodal67.259.3Gemini 3.7 Flash
Long context65.569.0Step 5 Preview

On one chart

Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.

2550 avg75ReasoningCodingAgents & toolsMathsKnowledgeLong contextMultimodalHumanpreferenceGemini 3.7 Flash, Reasoning: 67.8Gemini 3.7 Flash, Coding: 62.3Gemini 3.7 Flash, Agents & tools: 48.9Gemini 3.7 Flash, Maths: 62.6Gemini 3.7 Flash, Knowledge: 65.9Gemini 3.7 Flash, Long context: 65.5Gemini 3.7 Flash, Multimodal: 67.2Gemini 3.7 Flash, Human preference: 68.1Step 5 Preview, Reasoning: 51.7Step 5 Preview, Coding: 49.9Step 5 Preview, Agents & tools: 49.1Step 5 Preview, Maths: 60.1Step 5 Preview, Knowledge: 53.7Step 5 Preview, Long context: 69.0Step 5 Preview, Multimodal: 59.3Step 5 Preview, Human preference: 63.1

Not drawn, measured for fewer than two of these models: Instruction following.

Model, click to add or removeIndex

Where sources disagree

Across the 35 benchmarks both have, Gemini 3.7 Flash leads 20, Step 5 Preview leads 14 and 1 is a tie. Every overall index that rates both puts Gemini 3.7 Flash ahead.

Overall indexGemini 3.7 Flash (high)Step 5 PreviewLeader
BenchLeader Index62.156.0Gemini 3.7 Flash
Arena Text14891449Gemini 3.7 Flash

Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).

Price, speed and latency

List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.

MeasureGemini 3.7 Flash (high)Step 5 Preview
Blended price per 1M tokens$1.50$1.43
Input / output per 1M$0.750 / $3.75$1.00 / $2.70
Cached input per 1M$0.075$0.105
Price per task: chat reply$0.0014$0.0012
Price per task: summarize a 30-page report$0.011$0.014
Price per task: code edit$0.010$0.010
Price per task: agentic coding session$0.060$0.071
Price per task: structured extraction$0.0023$0.0025
Agentic coding session, input mostly cached$0.030$0.030
Cost per benchmark task (AA Index)–$2.35
Output tokens per task (AA Index)–113k
Of which reasoning–80k
Time per task (AA Index)–15 min
Output speed95 tok/s124 tok/s
Time to first answer2.2 s23.5 s
Time to a complete answer, last measured 13 Nov 2026–54.4 s

Each model is compared at its best-ranked reasoning-effort setting. Gemini 3.7 Flash (high) scores 62.1 and takes 2 s to its first answer token; Gemini 3.7 Flash (medium) scores 60.1 in 2 s, Gemini 3.7 Flash (low) scores 57.3 in 2 s. Step 5 Preview scores 56.0 and takes 24 s to its first answer token. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.

Every benchmark

135 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.

MetricPublisherGemini 3.7 Flash (high)Step 5 Preview
BenchLeader IndexBenchLeader62.156.0
Agents & tools scoreBenchLeader48.949.1
Coding scoreBenchLeader62.349.9
Human preference scoreBenchLeader68.163.1
Knowledge scoreBenchLeader65.953.7
Long context scoreBenchLeader65.569.0
Maths scoreBenchLeader62.660.1
Multimodal scoreBenchLeader67.259.3
Reasoning scoreBenchLeader67.851.7
Blended price $/MBenchLeader$1.50$1.43
Output speedBenchLeader95 tok/s124 tok/s
Time to first answerBenchLeader2.2 s23.5 s
Context windowBenchLeader1,048,576 tokens1,000,000 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub94.8%–
Arena Hard PromptsArena (LMArena)15101477
LiveBench 2.0 ReasoningLiveBench–83.8%†
LiveBench 1.0 ReasoningLiveBench87.8%–
GPQA Diamond (AA)Artificial Analysis94.5%–
Humanity's Last Exam (AA)Artificial Analysis47.9%46.5%
HLE-Diamond (AA)Artificial Analysis–24.3%
GPQA Diamond (Vals)Vals AI93.9%–
ARC-AGI-1ARC Prize95.5%–
ARC-AGI-2ARC Prize84.6%–
CritPtArtificial Analysis14.3%20.9%
CritPt v2 hard subset (AA)Artificial Analysis–32.5%
MysteryMechanismVals AI–18.5%
Arena ExpertArena (LMArena)15191489
Chess PuzzlesEpoch AI Benchmarking Hub66.0%–
Mystery Game PuzzlesEpoch AI Benchmarking Hub37.0%–
LMCAEpoch AI Benchmarking Hub50.4%–
DTBenchEpoch AI Benchmarking Hub96.8%–
ForecastBenchForecastBench66.2%–
Coding
SciCodeSciCode57.2%58.9%
FrontierCodeCognition43.6%†–
Arena CodingArena (LMArena)15201505
Arena WebDevArena (LMArena)15881573†
LiveBench 2.0 CodingLiveBench–67.6%†
LiveBench 1.0 CodingLiveBench78.9%–
Show all 135 benchmarks
MetricPublisherGemini 3.7 Flash (high)Step 5 Preview
Coding
LiveCodeBenchVals AI88.7%–
IOIVals AI67.8%38.1%
SWE-bench (Vals)Vals AI80.8%–
Code MigrationVals AI34.8%39.0%
Code Migration v1.1Vals AI–47.0%
ProgramBenchVals AI0.0%–
SWE-bench MultimodalVals AI–40.8%
Vibe Code Bench v1.1Vals AI70.4%70.4%
MirrorCodeEpoch AI Benchmarking Hub41.1%–
DeepSWE v1.1Epoch AI Benchmarking Hub65.3%–
ALE-BenchEpoch AI Benchmarking Hub904.3–
FrontierSWEEpoch AI Benchmarking Hub20.3%–
Terminal-Bench 4.0 (AA)Artificial Analysis13.6%33.3%
Terminal-Bench 2.1 (AA)Artificial Analysis85.8%–
Design Arena WebsiteDesign Arena1295†–
FrontierCode 1.1 (AA)Artificial Analysis–38.4%
FrontierCode 2.0 (AA)Artificial Analysis–12.0%
Agents & tools
Terminal-BenchTerminal-Bench12.4%–
Remote Labor IndexScale AI / CAIS5.0%†–
APEX-AgentsMercor67.8%†–
Arena AgentArena (LMArena)-2.2-3.7
LiveBench Long AgenticLiveBench–19.4%†
LiveBench BrowserLiveBench–24.8%†
LiveBench Agentic CodingLiveBench58.3%–
Terminal-Bench 2.1 (Vals)Vals AI77.5%–
GDPval-AA v2.1, normalised scoreArtificial Analysis–54.3%
GDPval-AA v2.1Artificial Analysis14041591
τ³-Banking (AA)Artificial Analysis32.8%–
ITBench SRE (AA)Artificial Analysis–55.6%
Analyst Agent (AA)Artificial Analysis60.0%35.0%
APEX-Agents (AA)Artificial Analysis–38.0%
CyberBenchVals AI43.8%–
Finance Agent v2Vals AI59.0%50.7%
Finance Agent v2.1Vals AI–58.9%
Harvey's Legal Agent Benchmark v1.1Vals AI–12.9%
Harvey's Legal Agent Benchmark v1Vals AI8.8%9.2%
Legal Research BenchVals AI34.6%40.4%
SkillsBenchVals AI65.9%–
SREBenchVals AI4.6%0.4%
Tax Agent BenchVals AI22.8%27.3%
Terminal-Bench 4.0 (Vals)Vals AI12.1%31.8%
Terminal-Bench ScienceVals AI7.1%2.9%
Arena Agent Task CompletionArena (LMArena)-3.5-1.5
AutomationBenchArtificial Analysis62.0%51.0%
GDP.pdf (AA)Artificial Analysis23.6%14.8%
EnterpriseOps-GymArtificial Analysis50.4%†47.2%
AA-Briefcase v1.1Artificial Analysis11151424
Design Arena AgenticDesign Arena1222†–
Terminal-Bench-Science 0.1 (AA)Artificial Analysis–2.4%
CWE-Bench (AA)Artificial Analysis–52.5%
DeepSec-Bench (AA)Artificial Analysis–15.8%
CyberGym end-to-end (AA)Artificial Analysis–74.8%
MKA (AA)Artificial Analysis–27.6%
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub71.6%–
FrontierMath Tier 4Epoch AI Benchmarking Hub36.6%–
OTIS Mock AIMEEpoch AI Benchmarking Hub97.2%–
ProofBenchVals AI58.0%†42.0%
LiveBench MathematicsLiveBench93.5%–
Arena MathsArena (LMArena)15021470
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub69.2%–
LiveBench 2.0 Data AnalysisLiveBench–63.5%†
LiveBench ResearchLiveBench–75.5%†
LiveBench 1.0 Data AnalysisLiveBench68.0%–
AA-Omniscience v1.1Artificial Analysis–16.4
MMLU-ProVals AI90.1%–
LegalBenchVals AI87.3%–
TaxEvalVals AI74.7%–
BioMysteryBenchVals AI–67.4%
Excel Modeling BenchmarkVals AI71.3%57.1%
Excel Modeling Benchmark v1.1Vals AI–69.5%
MedCodeVals AI53.4%–
MedScribeVals AI83.9%–
Public Benefits Bench v1.1Vals AI–58.1%
GDP.pdfEpoch AI Benchmarking Hub23.8%–
AA-Omniscience v1.1: accuracyArtificial Analysis–36.4%
AA-Omniscience v1.1: non-hallucinationArtificial Analysis–68.7%
Instruction following
LiveBench LanguageLiveBench85.5%–
LiveBench Instruction FollowingLiveBench79.9%–
Arena Instruction FollowingArena (LMArena)14861450
Human preference
Arena TextArena (LMArena)14891449
EQ-Bench Creative Writing v3EQ-Bench1720†–
Arena Creative WritingArena (LMArena)14871413
Arena Multi-turnArena (LMArena)14901445
Arena Longer QueriesArena (LMArena)15011461
Multimodal
Arena VisionArena (LMArena)13131263
LiveBench DesignLiveBench–70.8%†
MMMU-ProArtificial Analysis85.5%76.4%
MMMU-Pro (Vals)Vals AI89.0%–
MortgageTaxVals AI66.7%–
Humanity's Sixth SenseScale AI SEAL39.8%–
Furniture AssemblyEpoch AI Benchmarking Hub26.7%–
Code2Video-AAArtificial Analysis–821
AA-3DArtificial Analysis–859
Long context
AA-LCR v1.1Artificial Analysis81.7%88.3%
Arena DocumentArena (LMArena)1468–
MLCRArtificial Analysis12.2%18.3%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub158.4†–
LiveBench 2.0LiveBench–54.9%†
LiveBench 1.0LiveBench78.8%–
AA Intelligence Index v5.0.1Artificial Analysis–28.5
AA Intelligence Index v4.3.2Artificial Analysis39.143.7
AA Progress Index v1Artificial Analysis317.4318.2
Vals IndexVals AI–54.8
Vals Finance IndexVals AI–55.4%
Vals RSI IndexVals AI18.3%–
Vals Science IndexVals AI–29.6%
Vals Legal IndexVals AI–70.2%
AA Cyber IndexArtificial Analysis–47.7
Safety & honesty
DistressBenchScale AI SEAL69.0%†–
Metadata
AA Openness IndexArtificial Analysis–33.3

Data as of 2026-11-16. Best configuration of each model; every score links to its source on the model pages.

What changed

  • Step 5 Preview: BenchLeader Index 54.2 → 56.0.
  • Step 5 Preview: BenchLeader Index 54.7 → 54.2.
  • Gemini 3.7 Flash (high): 2 new results (GDPval-AA v2.1, AA-Briefcase v1.1).
  • Step 5 Preview: 1 score revision.
  • Gemini 3.7 Flash (high): BenchLeader Index 62.6 → 62.1.
  • Gemini 3.7 Flash (high): 1 score revision.
  • Step 5 Preview: 7 score revisions.
  • Gemini 3.7 Flash (high): 1 new result (Arena Document).
  • Step 5 Preview: 2 new results (Vals Index, Harvey's Legal Agent Benchmark v1.1).
  • Gemini 3.7 Flash (high): 1 new result (MirrorCode).

Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs

Gemini 3.7 Flash vs Step 5 Preview: questions

Is Gemini 3.7 Flash better than Step 5 Preview?
Gemini 3.7 Flash (high) leads on quality: 62.1 vs 56.0. The BenchLeader Index combines every independent quality benchmark; Gemini 3.7 Flash (high) is ahead overall as of 16 Nov 2026, but check the category scores for your use.
Gemini 3.7 Flash vs Step 5 Preview: which is cheaper, and will that change?
Step 5 Preview is cheaper today: $1.43 against $1.50 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is Gemini 3.7 Flash or Step 5 Preview better for coding?
Gemini 3.7 Flash scores higher in coding (62 vs 50 on the category index, where 50 is average). On SciCode, the top shared coding benchmark, it is 57.2% for Gemini 3.7 Flash against 58.9% for Step 5 Preview.
Which is faster, Gemini 3.7 Flash or Step 5 Preview?
Step 5 Preview streams ~1.5× faster (124 vs 95 output tokens per second, Artificial Analysis 7-day median).
Is Gemini 3.7 Flash or Step 5 Preview better for agents and tool use?
They are level in agentic tasks (49 vs 49 on the category index). On Arena Agent, the top shared agents & tools benchmark, it is -2.2 for Gemini 3.7 Flash against -3.7 for Step 5 Preview.
More questions (4)
Which is better for long documents?
Step 5 Preview scores higher on long-context benchmarks (69 vs 66). Both take about 1M tokens: Gemini 3.7 Flash 1,048,576, Step 5 Preview 1,000,000.
Which has lower latency?
Gemini 3.7 Flash answers first: 2.2 s against 23.5 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
Do independent benchmarks agree on Gemini 3.7 Flash vs Step 5 Preview?
Across the 35 benchmarks both have, Gemini 3.7 Flash leads 20, Step 5 Preview leads 14 and 1 is a tie. Every overall index that rates both puts Gemini 3.7 Flash ahead.
When was Step 5 Preview released?
Step 5 Preview was released on 18 Sep 2026.