Enter opens the highlighted result. Esc closes.

Claude Opus 5 vs Gemini 3.8 Flash

Claude Opus 5 leads Gemini 3.8 Flash on coding, 83.0 to 61.6 on the Vibe Coding Index, and Gemini 3.8 Flash costs 6.7x less per output token ($3.75 vs $25.00 per million). Updated October 21, 2026; 121 benchmarks have a result for both.

Replace Claude Opus 5 Replace Gemini 3.8 Flash
Claude Opus 5Anthropic
Coding
83.0
Overall
81.6
Output $/M
$25.00
101
Gemini 3.8 FlashGoogle
Coding
61.6
Overall
69.0
Output $/M
$3.75
18
vs boards won of 121 shared, 2 tied

Claude Opus 5 or Gemini 3.8 Flash: which to pick

Pick Claude Opus 5 for coding: it scores 83.0 on the Vibe Coding Index against 61.6 for Gemini 3.8 Flash.

Pick Gemini 3.8 Flash for output tokens 6.7x cheaper ($3.75 vs $25.00 per million) and a larger context window (1.05M against 1M tokens).

Model AClaude Opus 5Anthropic83.0/1006 independent benchmarks$1.33 per solved taskModel BGemini 3.8 FlashGoogle61.6/1006 independent benchmarks$2.74 per solved task
Shared results
121
Claude Opus 5 only
10
Gemini 3.8 Flash only
7
Like-for-like categories
8 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
Claude Opus 5 83.0 from 6 benchmarks, Gemini 3.8 Flash 61.6 from 6 benchmarks.
Widest category
Math, 20.6 points apart on average over 6 shared benchmarks.
Widest single gap
FrontierMath, tier 4: Claude Opus 5 73.2% against Gemini 3.8 Flash 22.0%, 51.2 points apart on a 0 to 100 scale.
Output price
Claude Opus 5 $25.00 against Gemini 3.8 Flash $3.75 per million output tokens, 6.7x apart.
Context window
Claude Opus 5 takes 1M tokens, Gemini 3.8 Flash 1.05M.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryClaude Opus 5BarsGemini 3.8 FlashTies
Coding232
Intelligence20
Agentic work2011 tie
Reasoning106
Math501 tie
Knowledge82
Web design90
General247
All shared boards101182 ties
CodingLike-for-likeSix of these boards feed the Vibe Coding Index232
Coding benchmarks for Claude Opus 5 and Gemini 3.8 Flash
BenchmarkClaude Opus 5Gemini 3.8 FlashLeader
LiveBench coding90.8%82.2%Claude Opus 5
LMArena WebDev16891585Claude Opus 5
AA Coding Index78.076.3Claude Opus 5
LiveCodeBench (Vals AI)89.0%89.5%Gemini 3.8 Flash
SWE-bench Verified (Vals AI)97.0%80.0%Claude Opus 5
Terminal-Bench 4 (Vals AI)53.5%19.7%Claude Opus 5
Vibe Code Bench (Vals AI)88.4%78.7%Claude Opus 5
Code migration (Vals AI)57.5%36.5%Claude Opus 5
ProgramBench (Vals AI)3.0%1.0%Claude Opus 5
IOI (Vals AI)84.3%56.9%Claude Opus 5
Terminal-Bench 2.1 (Vals AI)84.6%81.3%Claude Opus 5
SWE-bench Multimodal (Vals AI)Not measured40.8%Only Gemini 3.8 Flash measured
Vibe Code Bench 1-100 (Vals AI)28.5%18.8%Claude Opus 5
SkillsBench (Vals AI)60.4%58.0%Claude Opus 5
CyberBench (Vals AI)65.4%43.8%Claude Opus 5
SciCode56.4%56.6%Gemini 3.8 Flash
FrontierCode53.4%41.2%Claude Opus 5
FrontierSWE52.0%19.6%Claude Opus 5
WeirdML91.8%84.8%Claude Opus 5
ALE-Bench2164.61269.7Claude Opus 5
LMArena Text: coding15331508Claude Opus 5
LMArena Text: software and IT15271507Claude Opus 5
LMArena WebDev: React17111577Claude Opus 5
LMArena WebDev: HTML16801616Claude Opus 5
LMArena WebDev: from an image16621593Claude Opus 5
SciCode13.8%12.3%Claude Opus 5
Neither model has a result on the other 13 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for Claude Opus 5 and Gemini 3.8 Flash
BenchmarkClaude Opus 5Gemini 3.8 FlashLeader
AA Intelligence Index50.840.9Claude Opus 5
Epoch Capabilities Index163.9157.4Claude Opus 5
Agentic workLike-for-likePublic boards, shown as published, kept out of the index201
Agentic work benchmarks for Claude Opus 5 and Gemini 3.8 Flash
BenchmarkClaude Opus 5Gemini 3.8 FlashLeader
AA Agentic Index56.540.2Claude Opus 5
Design Arena: agents building web apps12661238Claude Opus 5
Design Arena: agents building full-stack apps13061244Claude Opus 5
Design Arena: agents building mobile apps13481236Claude Opus 5
Design Arena: agents building Android apps12541074Claude Opus 5
Design Arena: agents building Godot games12791273Claude Opus 5
Design Arena: agents building games1261Not measuredOnly Claude Opus 5 measured
Design Arena: PowerPoint slides12651177Claude Opus 5
Design Arena: spreadsheetsNot measured1056Only Gemini 3.8 Flash measured
LiveBench research88.9%79.1%Claude Opus 5
LiveBench browser84.0%64.0%Claude Opus 5
LiveBench long agentic62.1%30.1%Claude Opus 5
SREBench (Vals AI)12.2%11.1%Claude Opus 5
Finance Agent (Vals AI)58.6%61.4%Gemini 3.8 Flash
Tax Agent Bench (Vals AI)45.5%32.1%Claude Opus 5
Legal Research Bench (Vals AI)55.3%38.9%Claude Opus 5
Harvey's Legal Agent Benchmark (Vals AI)Not measured12.1%Only Gemini 3.8 Flash measured
Excel Modeling Benchmark (Vals AI)73.6%72.2%Claude Opus 5
BioMysteryBench (Vals AI)79.3%62.2%Claude Opus 5
Terminal-Bench-Science (Vals AI)27.1%8.6%Claude Opus 5
Vals Science Index47.9%35.8%Claude Opus 5
Time Horizon Index (Vals AI)18.8%18.8%Tie
CUA-bench (Vals AI)9.0%4.2%Claude Opus 5
APEX-Agents65.8%64.3%Claude Opus 5
Remote Labor IndexNot measured5.8%Only Gemini 3.8 Flash measured
OSWorld 2.031.4%Not measuredOnly Claude Opus 5 measured
BALROG63.4%Not measuredOnly Claude Opus 5 measured
PostTrainBench35.0%Not measuredOnly Claude Opus 5 measured
Vending-Bench 2$11,182$5,094Claude Opus 5
LMArena Search1242Not measuredOnly Claude Opus 5 measured
Neither model has a result on the other 11 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index106
Reasoning benchmarks for Claude Opus 5 and Gemini 3.8 Flash
BenchmarkClaude Opus 5Gemini 3.8 FlashLeader
GPQA Diamond93.4%94.4%Gemini 3.8 Flash
LiveBench reasoning91.2%89.3%Claude Opus 5
MysteryMechanism (Vals AI)37.4%36.5%Claude Opus 5
MMMU Pro (Vals AI)89.9%89.1%Claude Opus 5
GPQA Diamond (Artificial Analysis run)93.7%95.3%Gemini 3.8 Flash
Humanity's Last Exam (Artificial Analysis run)54.9%47.8%Claude Opus 5
GPQA Diamond (Epoch AI run)93.9%95.4%Gemini 3.8 Flash
ARC-AGI97.5%98.5%Gemini 3.8 Flash
ARC-AGI-290.4%89.2%Claude Opus 5
SimpleBench80.6%Not measuredOnly Claude Opus 5 measured
Chess puzzles43.0%58.0%Gemini 3.8 Flash
Mystery game puzzles59.0%47.0%Claude Opus 5
EBR-bench45.7%Not measuredOnly Claude Opus 5 measured
LMArena Text: hard prompts15271509Claude Opus 5
LMArena Text: expert15541526Claude Opus 5
LMArena Text: hard prompts in English15231504Claude Opus 5
LMArena Vision: diagrams13481345Claude Opus 5
LMArena Vision: homework13381351Gemini 3.8 Flash
Neither model has a result on the other 4 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index50
Math benchmarks for Claude Opus 5 and Gemini 3.8 Flash
BenchmarkClaude Opus 5Gemini 3.8 FlashLeader
OTIS Mock AIME98.9%98.9%Tie
FrontierMath, tiers 1 to 385.6%68.4%Claude Opus 5
FrontierMath, tier 473.2%22.0%Claude Opus 5
ProofBench99.0%48.0%Claude Opus 5
LMArena Text: math15331532Claude Opus 5
LMArena Text: mathematical work15441519Claude Opus 5
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index82
Knowledge benchmarks for Claude Opus 5 and Gemini 3.8 Flash
BenchmarkClaude Opus 5Gemini 3.8 FlashLeader
MMLU Pro91.6%90.2%Claude Opus 5
LegalBench (Vals AI)87.0%87.0%Gemini 3.8 Flash
CorpFin (Vals AI)73.2%Not measuredOnly Claude Opus 5 measured
TaxEval (Vals AI)75.1%74.5%Claude Opus 5
MortgageTax (Vals AI)72.1%65.3%Claude Opus 5
MedCode (Vals AI)63.6%48.1%Claude Opus 5
SimpleQA Verified59.9%69.7%Gemini 3.8 Flash
LMArena Text: science15311523Claude Opus 5
LMArena Text: medicine and healthcare15251508Claude Opus 5
LMArena Text: legal and government15301497Claude Opus 5
LMArena Text: business and finance15011479Claude Opus 5
FACTS Benchmark SuiteNot measured69.4%Only Gemini 3.8 Flash measured
Neither model has a result on the other 6 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index90
Web design benchmarks for Claude Opus 5 and Gemini 3.8 Flash
BenchmarkClaude Opus 5Gemini 3.8 FlashLeader
Design Arena: websites13091300Claude Opus 5
Design Arena: UI components13491319Claude Opus 5
Design Arena: game dev13381305Claude Opus 5
Design Arena: data viz13291269Claude Opus 5
Design Arena: 3D13351295Claude Opus 5
Design Arena: all code categories13241304Claude Opus 5
Design Arena: SVG13111295Claude Opus 5
Design Arena: ASCII art13521265Claude Opus 5
LiveBench design83.0%72.7%Claude Opus 5
GeneralLike-for-likePublic boards, shown as published, kept out of the index247
General benchmarks for Claude Opus 5 and Gemini 3.8 Flash
BenchmarkClaude Opus 5Gemini 3.8 FlashLeader
Vals IndexNot measured56.0%Only Gemini 3.8 Flash measured
LiveBench data analysis74.6%54.0%Claude Opus 5
Public Benefits Bench (Vals AI)76.9%65.3%Claude Opus 5
SAGE (Vals AI)49.4%35.1%Claude Opus 5
MedScribe (Vals AI)91.0%84.5%Claude Opus 5
Vals Multimodal Index73.9%Not measuredOnly Claude Opus 5 measured
Long context reasoning (AA-LCR)82.0%84.0%Gemini 3.8 Flash
DTBench97.9%95.7%Claude Opus 5
LMCA64.5%52.9%Claude Opus 5
Surface Evolver BenchNot measured76.9%Only Gemini 3.8 Flash measured
Furniture assembly60.8%31.7%Claude Opus 5
LMArena Text15091499Claude Opus 5
LMArena Text: instruction following15181492Claude Opus 5
LMArena Text: creative writing14931488Claude Opus 5
LMArena Vision13201315Claude Opus 5
LMArena Text: English15081498Claude Opus 5
LMArena Text: non-English15021491Claude Opus 5
LMArena Text: Chinese15731553Claude Opus 5
LMArena Text: German15271501Claude Opus 5
LMArena Text: Japanese15201507Claude Opus 5
LMArena Text: Korean15171464Claude Opus 5
LMArena Text: French15191496Claude Opus 5
LMArena Text: Spanish15231487Claude Opus 5
LMArena Text: Russian15071514Gemini 3.8 Flash
LMArena Text: Polish15221497Claude Opus 5
LMArena Text: multi-turn15001503Gemini 3.8 Flash
LMArena Text: longer queries15171507Claude Opus 5
LMArena Text: writing and language14991494Claude Opus 5
LMArena Text: entertainment and media14811480Claude Opus 5
LMArena Vision: English13151318Gemini 3.8 Flash
LMArena Vision: Chinese14001416Gemini 3.8 Flash
LMArena Vision: reading text13291333Gemini 3.8 Flash
LMArena Vision: humor12931321Gemini 3.8 Flash
LMArena Vision: creative writing13131299Claude Opus 5
LMArena Document1516Not measuredOnly Claude Opus 5 measured
Neither model has a result on the other 11 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Claude Opus 5 and Gemini 3.8 Flash
BenchmarkClaude Opus 5Gemini 3.8 FlashLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

Claude Opus 5$0.018Fits in one request
Gemini 3.8 Flash$0.0026Fits in one request

Gemini 3.8 Flash costs less

Repository review

50K input + 3K output tokens

Claude Opus 5$0.325Fits in one request
Gemini 3.8 Flash$0.049Fits in one request

Gemini 3.8 Flash costs less

Agent loop

200K input + 10K output tokens

Claude Opus 5$1.25Fits in one request
Gemini 3.8 Flash$0.188Fits in one request

Gemini 3.8 Flash costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of Claude Opus 5 and Gemini 3.8 Flash
SpecificationAnthropic Claude Opus 5Google Gemini 3.8 Flash
MakerAnthropicGoogle
WeightsNo public weights listedNo public weights listed
ReleasedJuly 24, 2026September 2, 2026
Context window1M1.05M
Input $/Mtok$5.00$0.75
Output $/Mtok$25.00$3.75
Coding index83.061.6
Independent measurements behind it66
Cost per solved task$1.33$2.74
SourceOpenRouter listingOpenRouter listing

Questions

Is Claude Opus 5 better than Gemini 3.8 Flash for coding?

Pick Claude Opus 5 for coding: it scores 83.0 on the Vibe Coding Index against 61.6 for Gemini 3.8 Flash. Pick Gemini 3.8 Flash for output tokens 6.7x cheaper ($3.75 vs $25.00 per million) and a larger context window (1.05M against 1M tokens).

Which is cheaper, Claude Opus 5 or Gemini 3.8 Flash?

Claude Opus 5 costs $5.00 per million input tokens and $25.00 per million output tokens; Gemini 3.8 Flash costs $0.75 and $3.75. Per solved SWE-bench Verified task, Claude Opus 5 costs $1.33 and Gemini 3.8 Flash $2.74.

Which has the bigger context window, Claude Opus 5 or Gemini 3.8 Flash?

Gemini 3.8 Flash, with 1.05M tokens against 1M.

How many benchmarks is the Claude Opus 5 vs Gemini 3.8 Flash comparison based on?

121 benchmarks have a result for both models as of October 21, 2026. Claude Opus 5's coding index rests on 6 independent benchmarks and Gemini 3.8 Flash's on 6. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/claude-opus-5-vs-gemini-3-8-flash" title="Claude Opus 5 vs Gemini 3.8 Flash, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/claude-opus-5-vs-gemini-3-8-flash">LLMBenchmark.ai, Claude Opus 5 vs Gemini 3.8 Flash</a></p>