BenchLeader

Gemini 3.5 Flash vs Step 5 Preview

Full profiles: Gemini 3.5 Flash (Google) and Step 5 Preview (StepFun): every benchmark, price and rank.

Updated · Data from 11 sources

As of 16 Nov 2026, Gemini 3.5 Flash is ahead of Step 5 Preview on quality (62.5 vs 56.0 on the BenchLeader Index), but Step 5 Preview is 2.4× cheaper at $1.43 vs $3.38 per 1M tokens blended. Choose Gemini 3.5 Flash for agents & tools (63 vs 49), coding (60 vs 50), and knowledge (63 vs 54) and fast first answers (9.9 s vs 23.5 s), where it also streams faster; choose Step 5 Preview for long context (69 vs 61) and maths (60 vs 58).

Key facts (as of 16 Nov 2026)

  • Gemini 3.5 Flash scores 62.5 on the BenchLeader Index and Step 5 Preview scores 56.0 (BenchLeader, 16 Nov 2026).
  • Gemini 3.5 Flash costs $3.38 per 1M tokens blended (list price, 16 Nov 2026).
  • Step 5 Preview costs $1.43 per 1M tokens blended (list price, 16 Nov 2026).
  • Gemini 3.5 Flash streams faster than Step 5 Preview (Artificial Analysis, 7-day median, 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 3.5 Flash60 vs 50 category score; Arena Coding 1508 vs 1505High
Agents & tool useGemini 3.5 Flash63 vs 49 category score; Analyst Agent (AA) 45.0% vs 35.0%High
MathsStep 5 Preview60 vs 58 category score; ProofBench 42.0% vs 31.0%Medium
ReasoningGemini 3.5 Flash61 vs 52 category score; Arena Hard Prompts 1496 vs 1477High
KnowledgeGemini 3.5 Flash63 vs 54 category score; Excel Modeling Benchmark 63.5% vs 57.1%High
Long documentsStep 5 Preview69 vs 61 category score; AA-LCR v1.1 88.3% vs 73.3%High
Images and other mediaGemini 3.5 Flash66 vs 59 category score; Arena Vision 1308 vs 1263Medium
Chat quality (human preference)Gemini 3.5 Flash67 vs 63 on Arena votesMedium
Lowest costStep 5 Preview$1.43 vs $3.38 per 1M blended; agentic coding session $0.071 vs $0.126–
Fastest first answerGemini 3.5 Flash9.9 s (Gemini 3.5 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 3.5 Flash (high) leads on quality: 62.5 vs 56.0.
  • Gemini 3.5 Flash (high) is stronger in agents & tools, coding, human preference, knowledge, multimodal, reasoning.
  • Step 5 Preview is stronger in long context, maths.
  • Step 5 Preview is 2.4× cheaper ($1.43 vs $3.38 per 1M blended).
  • Gemini 3.5 Flash (high) streams 1.4× faster (176 vs 124 tokens per second).

At a glance

SpecGemini 3.5 FlashStep 5 Preview
DeveloperGoogleStepFun
Released19 May 202618 Sep 2026
Context window (tokens)1,048,5761,000,000
Compared atGemini 3.5 Flash (high)Step 5 Preview
Open weightsNoYes
Price in / out / cached input per 1M$1.50 / $9.00 / $0.150$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 3.5 Flash (high)Step 5 PreviewLeader
Reasoning60.651.7Gemini 3.5 Flash
Coding59.549.9Gemini 3.5 Flash
Agents & tools62.849.1Gemini 3.5 Flash
Maths57.760.1Step 5 Preview
Knowledge63.353.7Gemini 3.5 Flash
Instruction following74.4–not comparable
Human preference66.863.1Gemini 3.5 Flash
Multimodal66.259.3Gemini 3.5 Flash
Long context61.269.0Step 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 3.5 Flash, Reasoning: 60.6Gemini 3.5 Flash, Coding: 59.5Gemini 3.5 Flash, Agents & tools: 62.8Gemini 3.5 Flash, Maths: 57.7Gemini 3.5 Flash, Knowledge: 63.3Gemini 3.5 Flash, Long context: 61.2Gemini 3.5 Flash, Multimodal: 66.2Gemini 3.5 Flash, Human preference: 66.8Step 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 36 benchmarks both have, Gemini 3.5 Flash leads 18, Step 5 Preview leads 18. Every overall index that rates both puts Gemini 3.5 Flash ahead.

Overall indexGemini 3.5 Flash (high)Step 5 PreviewLeader
BenchLeader Index62.556.0Gemini 3.5 Flash
Arena Text14781449Gemini 3.5 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 3.5 Flash (high)Step 5 Preview
Blended price per 1M tokens$3.38$1.43
Input / output per 1M$1.50 / $9.00$1.00 / $2.70
Cached input per 1M$0.150$0.105
Price per task: chat reply$0.0033$0.0012
Price per task: summarize a 30-page report$0.023$0.014
Price per task: code edit$0.022$0.010
Price per task: agentic coding session$0.126$0.071
Price per task: structured extraction$0.0048$0.0025
Agentic coding session, input mostly cached$0.065$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 speed176 tok/s124 tok/s
Time to first answer9.9 s23.5 s
Time to a complete answer, last measured 13 Nov 202627.6 s54.4 s

Each model is compared at its best-ranked reasoning-effort setting. Gemini 3.5 Flash (high) scores 62.5 and takes 10 s to its first answer token; Gemini 3.5 Flash (medium) scores 61.8 in 9 s, Gemini 3.5 Flash (minimal) scores 55.6 in 3 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

153 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 3.5 Flash (high)Step 5 Preview
BenchLeader IndexBenchLeader62.556.0
Agents & tools scoreBenchLeader62.849.1
Coding scoreBenchLeader59.549.9
Human preference scoreBenchLeader66.863.1
Instruction following scoreBenchLeader74.4–
Knowledge scoreBenchLeader63.353.7
Long context scoreBenchLeader61.269.0
Maths scoreBenchLeader57.760.1
Multimodal scoreBenchLeader66.259.3
Reasoning scoreBenchLeader60.651.7
Blended price $/MBenchLeader$3.38$1.43
Output speedBenchLeader176 tok/s124 tok/s
Time to first answerBenchLeader9.9 s23.5 s
Context windowBenchLeader1,048,576 tokens1,000,000 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub92.8%–
SimpleBenchSimpleBench76.7%†–
Arena Hard PromptsArena (LMArena)14961477
LiveBench 2.0 ReasoningLiveBench–83.8%†
LiveBench 1.0 ReasoningLiveBench82.0%–
GPQA Diamond (AA)Artificial Analysis92.2%–
Humanity's Last Exam (AA)Artificial Analysis42.7%46.5%
HLE-Diamond (AA)Artificial Analysis–24.3%
GPQA Diamond (Vals)Vals AI92.7%–
EnigmaEvalScale AI SEAL25.4%–
ARC-AGI-1ARC Prize92.5%–
ARC-AGI-2ARC Prize72.1%–
CritPtArtificial Analysis13.1%20.9%
CritPt v2 hard subset (AA)Artificial Analysis–32.5%
MysteryMechanismVals AI–18.5%
Arena ExpertArena (LMArena)14911489
Chess PuzzlesEpoch AI Benchmarking Hub45.0%–
Mystery Game PuzzlesEpoch AI Benchmarking Hub32.0%–
LMCAEpoch AI Benchmarking Hub47.1%–
DTBenchEpoch AI Benchmarking Hub94.7%–
ForecastBenchForecastBench64.0%†–
Coding
SWE-bench Verified (Epoch)Epoch AI Benchmarking Hub79.3%–
SciCodeSciCode53.9%58.9%
WeirdMLWeirdML62.6%–
Arena CodingArena (LMArena)15081505
Show all 153 benchmarks
MetricPublisherGemini 3.5 Flash (high)Step 5 Preview
Coding
Arena WebDevArena (LMArena)15001573†
LiveBench 2.0 CodingLiveBench–67.6%†
LiveBench 1.0 CodingLiveBench78.2%–
LiveCodeBenchVals AI87.6%–
IOIVals AI–38.1%
SWE-bench (Vals)Vals AI78.8%–
Code MigrationVals AI26.8%39.0%
Code Migration v1.1Vals AI–47.0%
ProgramBenchVals AI0.0%–
SWE-bench MultimodalVals AI–40.8%
Vibe Code Bench v1.1Vals AI48.7%70.4%
Surface Evolver BenchEpoch AI Benchmarking Hub58.1%†–
DeepSWE v1.1Epoch AI Benchmarking Hub36.1%–
ALE-BenchEpoch AI Benchmarking Hub911.0–
Terminal-Bench 4.0 (AA)Artificial Analysis6.6%33.3%
Terminal-Bench 2.1 (AA)Artificial Analysis78.7%–
GameBench 2 OverallDuelLab GameBench 249.4†–
GameBench 2 by reasoning settingDuelLab GameBench 263.2–
Design Arena WebsiteDesign Arena1260†–
FrontierCode 1.1 (AA)Artificial Analysis–38.4%
FrontierCode 2.0 (AA)Artificial Analysis–12.0%
Agents & tools
APEX-AgentsMercor27.5%†–
Arena AgentArena (LMArena)–-3.7
LiveBench Long AgenticLiveBench–19.4%†
LiveBench BrowserLiveBench–24.8%†
LiveBench Agentic CodingLiveBench49.0%–
Terminal-Bench HardArtificial Analysis40.9%–
τ²-Bench Telecom (AA)Artificial Analysis95.3%–
Terminal-Bench 2.1 (Vals)Vals AI74.2%–
MCP AtlasScale AI SEAL83.6%–
HiL-BenchScale AI SEAL27.7%†–
GDPval-AA v2.1, normalised scoreArtificial Analysis–54.3%
GDPval-AA v2.1Artificial Analysis12111591
τ³-Banking (AA)Artificial Analysis32.2%–
ITBench SRE (AA)Artificial Analysis40.4%55.6%
Analyst Agent (AA)Artificial Analysis45.0%35.0%
APEX-Agents (AA)Artificial Analysis47.0%38.0%
Finance Agent v2Vals AI57.9%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 AI2.5%9.2%
Legal Research BenchVals AI30.8%40.4%
SkillsBenchVals AI52.7%–
SREBenchVals AI–0.4%
Tax Agent BenchVals AI21.6%27.3%
Terminal-Bench 2.0 (Vals)Vals AI67.4%–
Terminal-Bench 4.0 (Vals)Vals AI6.1%31.8%
Terminal-Bench ScienceVals AI5.7%2.9%
Web Search IndexVals AI37.0%–
Arena Agent Task CompletionArena (LMArena)–-1.5
GBAEvalEpoch AI Benchmarking Hub6.7%†–
Vending-Bench 2Andon Labs5396.4†–
AutomationBenchArtificial Analysis42.1%51.0%
GDP.pdf (AA)Artificial Analysis19.8%14.8%
EnterpriseOps-GymArtificial Analysis50.1%47.2%
AA-Briefcase v1.1Artificial Analysis8761424
Design Arena AgenticDesign Arena1174†–
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 Hub62.8%–
FrontierMath Tier 4Epoch AI Benchmarking Hub26.8%–
OTIS Mock AIMEEpoch AI Benchmarking Hub95.6%–
ProofBenchVals AI31.0%42.0%
LiveBench MathematicsLiveBench88.2%–
AIME 2026MathArena95.0%–
HMMT February 2026MathArena95.5%–
MathArena ApexMathArena32.3%–
Arena MathsArena (LMArena)14971470
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub66.2%–
LiveBench 2.0 Data AnalysisLiveBench–63.5%†
LiveBench ResearchLiveBench–75.5%†
LiveBench 1.0 Data AnalysisLiveBench64.9%–
AA-Omniscience v1.1Artificial Analysis–16.4
MMLU-ProVals AI89.5%–
LegalBenchVals AI83.6%–
CorpFinVals AI64.7%–
TaxEvalVals AI74.4%–
BioMysteryBenchVals AI–67.4%
Excel Modeling BenchmarkVals AI63.5%57.1%
Excel Modeling Benchmark v1.1Vals AI–69.5%
MedCodeVals AI55.8%–
MedScribeVals AI76.6%–
Public Benefits Bench v1.1Vals AI59.5%58.1%
Financial Coaching BenchVals AI63.0%–
DrugDiscoveryBenchScale AI SEAL50.0%–
EBR-benchEpoch AI Benchmarking Hub4.8%–
GDP.pdfEpoch AI Benchmarking Hub14.0%†–
AA-Omniscience v1.1: accuracyArtificial Analysis–36.4%
AA-Omniscience v1.1: non-hallucinationArtificial Analysis–68.7%
Instruction following
LiveBench LanguageLiveBench84.6%–
LiveBench Instruction FollowingLiveBench75.6%–
IFBenchArtificial Analysis76.3%–
Arena Instruction FollowingArena (LMArena)14651450
Human preference
Arena TextArena (LMArena)14781449
EQ-Bench 4EQ-Bench1077†–
Arena Creative WritingArena (LMArena)14651413
Arena Multi-turnArena (LMArena)14801445
Arena Longer QueriesArena (LMArena)14831461
Multimodal
Arena VisionArena (LMArena)13081263
LiveBench DesignLiveBench–70.8%†
MMMU-ProArtificial Analysis84.3%76.4%
MMMU-Pro (Vals)Vals AI88.3%–
MortgageTaxVals AI68.1%–
Humanity's Sixth SenseScale AI SEAL32.6%–
GeoBenchGeoBench85.0%†–
Blueprint-Bench 2Epoch AI Benchmarking Hub33.6%†–
Code2Video-AAArtificial Analysis–821
AA-3DArtificial Analysis–859
Long context
AA-LCR v1.1Artificial Analysis73.3%88.3%
Arena DocumentArena (LMArena)1455–
MLCRArtificial Analysis10.0%18.3%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub155.3†–
LiveBench 2.0LiveBench–54.9%†
LiveBench 1.0LiveBench74.6%–
AA Intelligence Index v5.0.1Artificial Analysis–28.5
AA Intelligence Index v4.3.2Artificial Analysis32.643.7
AA Progress Index v1Artificial Analysis305.1318.2
Vals IndexVals AI–54.8
Vals Finance IndexVals AI–55.4%
Vals Science IndexVals AI–29.6%
Vals Multimodal IndexVals AI62.9%–
Vals Legal IndexVals AI–70.2%
AA Cyber IndexArtificial Analysis–47.7
Safety & honesty
DistressBenchScale AI SEAL66.9%†–
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.

DuelLab GameBench 2, release gb2-20261111-7b598bb13de4 (https://benchmarks.duellab.org/releases/gb2-20261111-7b598bb13de4/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.

What changed

  • Step 5 Preview: BenchLeader Index 54.2 → 56.0.
  • Step 5 Preview: BenchLeader Index 54.7 → 54.2.
  • Gemini 3.5 Flash (high): 1 new result (GDPval-AA v2.1); 1 score revision.
  • Gemini 3.5 Flash (high): 3 new results (AIME 2026, HMMT February 2026, MathArena Apex).
  • Step 5 Preview: 1 score revision.
  • Step 5 Preview: 7 score revisions.
  • Gemini 3.5 Flash (high): 1 score revision.
  • Step 5 Preview: 2 new results (Vals Index, Harvey's Legal Agent Benchmark v1.1).
  • Step 5 Preview: 1 score revision.
  • Gemini 3.5 Flash (high): BenchLeader Index 63.2 → 62.6.

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

Gemini 3.5 Flash vs Step 5 Preview: questions

Is Gemini 3.5 Flash better than Step 5 Preview?
Gemini 3.5 Flash (high) leads on quality: 62.5 vs 56.0. The BenchLeader Index combines every independent quality benchmark; Gemini 3.5 Flash (high) is ahead overall as of 16 Nov 2026, but check the category scores for your use.
Gemini 3.5 Flash vs Step 5 Preview: which is cheaper, and will that change?
Step 5 Preview is cheaper today: $1.43 against $3.38 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is Gemini 3.5 Flash or Step 5 Preview better for coding?
Gemini 3.5 Flash scores higher in coding (60 vs 50 on the category index, where 50 is average). On SciCode, the top shared coding benchmark, it is 53.9% for Gemini 3.5 Flash against 58.9% for Step 5 Preview.
Which is faster, Gemini 3.5 Flash or Step 5 Preview?
Gemini 3.5 Flash streams faster (176 vs 124 output tokens per second, Artificial Analysis 7-day median).
Is Gemini 3.5 Flash or Step 5 Preview better for agents and tool use?
Gemini 3.5 Flash scores higher in agentic tasks (63 vs 49 on the category index, where 50 is average).
More questions (4)
Which is better for long documents?
Step 5 Preview scores higher on long-context benchmarks (69 vs 61). Both take about 1M tokens: Gemini 3.5 Flash 1,048,576, Step 5 Preview 1,000,000.
Which has lower latency?
Gemini 3.5 Flash answers first: 9.9 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 3.5 Flash vs Step 5 Preview?
Across the 36 benchmarks both have, Gemini 3.5 Flash leads 18, Step 5 Preview leads 18. Every overall index that rates both puts Gemini 3.5 Flash ahead.
When was Step 5 Preview released?
Step 5 Preview was released on 18 Sep 2026.