LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

Claude Opus 4.7 vs Gemma 4 26B A4B

Updated October 21, 2026. 51 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace Claude Opus 4.7 Replace Gemma 4 26B A4B
Claude Opus 4.7Anthropic
Coding
51.1
Overall
40.7
Output $/M
$25.00
50
Gemma 4 26B A4BGoogle
Coding
22.8
Overall
16.7
Output $/M
$0.255
1
vs boards won of 51 shared

Decision reading

Pick Claude Opus 4.7 for coding: it scores 51.1 on the Vibe Coding Index against 22.8 for Gemma 4 26B A4B.

Pick Gemma 4 26B A4B when output tokens cost 98.0x less, or when the benchmark that matters is IFBench, where it leads.

Model AClaude Opus 4.7Anthropic51.1/1001 independent benchmark$2.95 per solved taskModel BGemma 4 26B A4BGoogle22.8/1001 independent benchmarkNo cost per solved task
Shared results
51
Claude Opus 4.7 only
50
Gemma 4 26B A4B only
3
Like-for-like categories
7 of 9

Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.

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 4.7 51.1 from 1 benchmark, Gemma 4 26B A4B 22.8 from 1 benchmark.
Widest category
Agentic work, 40.4 points apart on average over 2 shared benchmarks.
Widest single gap
τ²-Bench Telecom: Claude Opus 4.7 88.6% against Gemma 4 26B A4B 43.6%, 45.0 points apart on a 0 to 100 scale.
Output price
Claude Opus 4.7 $25.00 against Gemma 4 26B A4B $0.255 per million output tokens, 98.0x apart.
Context window
Claude Opus 4.7 takes 1M tokens, Gemma 4 26B A4B 262K.

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 4.7BarsGemma 4 26B A4BTies
Coding80
Intelligence20
Agentic work20
Reasoning80
Math30
Knowledge40
General231
All shared boards501
CodingLike-for-likeFour of these boards feed the Vibe Coding Index80
Coding benchmarks for Claude Opus 4.7 and Gemma 4 26B A4B
BenchmarkClaude Opus 4.7Gemma 4 26B A4BLeader
LMArena WebDev15581360Claude Opus 4.7
AA Coding Index73.639.3Claude Opus 4.7
LiveCodeBench (Vals AI)85.1%Not measuredOnly Claude Opus 4.7 measured
SWE-bench Verified (Vals AI)82.0%Not measuredOnly Claude Opus 4.7 measured
Vibe Code Bench (Vals AI)71.0%Not measuredOnly Claude Opus 4.7 measured
Code migration (Vals AI)43.9%Not measuredOnly Claude Opus 4.7 measured
ProgramBench (Vals AI)0.0%Not measuredOnly Claude Opus 4.7 measured
Terminal-Bench 2.1 (Vals AI)68.5%Not measuredOnly Claude Opus 4.7 measured
SciCodeNot measured40.0%Only Gemma 4 26B A4B measured
Terminal-Bench Hard (Artificial Analysis run)54.5%25.0%Claude Opus 4.7
Terminal-Bench80.2%Not measuredOnly Claude Opus 4.7 measured
FrontierCode38.5%Not measuredOnly Claude Opus 4.7 measured
GSO-Bench44.1%Not measuredOnly Claude Opus 4.7 measured
WeirdML76.4%35.2%Claude Opus 4.7
MirrorCode31.1%Not measuredOnly Claude Opus 4.7 measured
ALE-BenchNot measured927.2Only Gemma 4 26B A4B measured
LMArena Text: coding15191446Claude Opus 4.7
LMArena Text: software and IT15161454Claude Opus 4.7
LMArena WebDev: React15601351Claude Opus 4.7
LMArena WebDev: HTML15541362Claude Opus 4.7
LMArena WebDev: from an image1577Not measuredOnly Claude Opus 4.7 measured
Neither model has a result on the other 17 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for Claude Opus 4.7 and Gemma 4 26B A4B
BenchmarkClaude Opus 4.7Gemma 4 26B A4BLeader
AA Intelligence Index40.716.7Claude Opus 4.7
Epoch Capabilities Index156.8141.9Claude Opus 4.7
Agentic workLike-for-likePublic boards, shown as published, kept out of the index20
Agentic work benchmarks for Claude Opus 4.7 and Gemma 4 26B A4B
BenchmarkClaude Opus 4.7Gemma 4 26B A4BLeader
AA Agentic Index38.62.9Claude Opus 4.7
Finance Agent (Vals AI)51.5%Not measuredOnly Claude Opus 4.7 measured
Tax Agent Bench (Vals AI)23.7%Not measuredOnly Claude Opus 4.7 measured
Legal Research Bench (Vals AI)38.5%Not measuredOnly Claude Opus 4.7 measured
Excel Modeling Benchmark (Vals AI)63.7%Not measuredOnly Claude Opus 4.7 measured
τ²-Bench Telecom88.6%43.6%Claude Opus 4.7
APEX-Agents49.2%Not measuredOnly Claude Opus 4.7 measured
OSWorld 2.018.2%Not measuredOnly Claude Opus 4.7 measured
PostTrainBench28.6%Not measuredOnly Claude Opus 4.7 measured
Vending-Bench 2$10,937Not measuredOnly Claude Opus 4.7 measured
LMArena Search1230Not measuredOnly Claude Opus 4.7 measured
Neither model has a result on the other 28 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index80
Reasoning benchmarks for Claude Opus 4.7 and Gemma 4 26B A4B
BenchmarkClaude Opus 4.7Gemma 4 26B A4BLeader
GPQA Diamond90.2%Not measuredOnly Claude Opus 4.7 measured
MMMU Pro (Vals AI)85.5%Not measuredOnly Claude Opus 4.7 measured
GPQA Diamond (Artificial Analysis run)91.4%79.2%Claude Opus 4.7
Humanity's Last Exam (Artificial Analysis run)42.3%19.3%Claude Opus 4.7
GPQA Diamond (Epoch AI run)90.2%73.2%Claude Opus 4.7
ARC-AGI93.5%Not measuredOnly Claude Opus 4.7 measured
ARC-AGI-275.8%Not measuredOnly Claude Opus 4.7 measured
SimpleBench61.7%Not measuredOnly Claude Opus 4.7 measured
Chess puzzles41.0%Not measuredOnly Claude Opus 4.7 measured
Mystery game puzzles28.0%Not measuredOnly Claude Opus 4.7 measured
EBR-bench19.0%Not measuredOnly Claude Opus 4.7 measured
LMArena Text: hard prompts15061438Claude Opus 4.7
LMArena Text: expert15211447Claude Opus 4.7
LMArena Text: hard prompts in English15081448Claude Opus 4.7
LMArena Vision: diagrams13361264Claude Opus 4.7
LMArena Vision: homework13371293Claude Opus 4.7
Neither model has a result on the other 5 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index30
Math benchmarks for Claude Opus 4.7 and Gemma 4 26B A4B
BenchmarkClaude Opus 4.7Gemma 4 26B A4BLeader
AIME96.3%Not measuredOnly Claude Opus 4.7 measured
OTIS Mock AIME97.8%82.2%Claude Opus 4.7
FrontierMath, tiers 1 to 370.2%Not measuredOnly Claude Opus 4.7 measured
FrontierMath, tier 431.7%Not measuredOnly Claude Opus 4.7 measured
ProofBench54.0%Not measuredOnly Claude Opus 4.7 measured
LMArena Text: math14981470Claude Opus 4.7
LMArena Text: mathematical work15091475Claude Opus 4.7
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index40
Knowledge benchmarks for Claude Opus 4.7 and Gemma 4 26B A4B
BenchmarkClaude Opus 4.7Gemma 4 26B A4BLeader
MMLU Pro89.9%Not measuredOnly Claude Opus 4.7 measured
LegalBench (Vals AI)85.3%Not measuredOnly Claude Opus 4.7 measured
CaseLaw (Vals AI)68.4%Not measuredOnly Claude Opus 4.7 measured
CorpFin (Vals AI)66.1%Not measuredOnly Claude Opus 4.7 measured
TaxEval (Vals AI)75.3%Not measuredOnly Claude Opus 4.7 measured
MortgageTax (Vals AI)70.3%Not measuredOnly Claude Opus 4.7 measured
MedCode (Vals AI)54.9%Not measuredOnly Claude Opus 4.7 measured
SimpleQA Verified51.7%Not measuredOnly Claude Opus 4.7 measured
LMArena Text: science15081446Claude Opus 4.7
LMArena Text: medicine and healthcare14951428Claude Opus 4.7
LMArena Text: legal and government15041457Claude Opus 4.7
LMArena Text: business and finance14901429Claude Opus 4.7
Neither model has a result on the other 5 boards here
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Web design benchmarks for Claude Opus 4.7 and Gemma 4 26B A4B
BenchmarkClaude Opus 4.7Gemma 4 26B A4BLeader
Neither model has a result on 9 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index231
General benchmarks for Claude Opus 4.7 and Gemma 4 26B A4B
BenchmarkClaude Opus 4.7Gemma 4 26B A4BLeader
SAGE (Vals AI)56.1%Not measuredOnly Claude Opus 4.7 measured
MedScribe (Vals AI)83.0%Not measuredOnly Claude Opus 4.7 measured
Vals Multimodal Index67.4%Not measuredOnly Claude Opus 4.7 measured
IFBench58.6%72.4%Gemma 4 26B A4B
Long context reasoning (AA-LCR)78.7%65.7%Claude Opus 4.7
DTBench94.7%74.9%Claude Opus 4.7
LMCA52.2%29.7%Claude Opus 4.7
Furniture assembly33.3%Not measuredOnly Claude Opus 4.7 measured
ExploitBench26.5%Not measuredOnly Claude Opus 4.7 measured
LMArena Text14901434Claude Opus 4.7
LMArena Text: instruction following14981420Claude Opus 4.7
LMArena Text: creative writing14861400Claude Opus 4.7
LMArena Vision13171260Claude Opus 4.7
LMArena Text: English14961444Claude Opus 4.7
LMArena Text: non-English14801421Claude Opus 4.7
LMArena Text: Chinese15311495Claude Opus 4.7
LMArena Text: German1495Not measuredOnly Claude Opus 4.7 measured
LMArena Text: Japanese1474Not measuredOnly Claude Opus 4.7 measured
LMArena Text: Korean1464Not measuredOnly Claude Opus 4.7 measured
LMArena Text: French15031460Claude Opus 4.7
LMArena Text: Spanish14941415Claude Opus 4.7
LMArena Text: Russian14941434Claude Opus 4.7
LMArena Text: Polish1489Not measuredOnly Claude Opus 4.7 measured
LMArena Text: multi-turn15051441Claude Opus 4.7
LMArena Text: longer queries15051427Claude Opus 4.7
LMArena Text: writing and language14921405Claude Opus 4.7
LMArena Text: entertainment and media14741403Claude Opus 4.7
LMArena Vision: English13221256Claude Opus 4.7
LMArena Vision: Chinese13711312Claude Opus 4.7
LMArena Vision: reading text13271268Claude Opus 4.7
LMArena Vision: humor13091272Claude Opus 4.7
LMArena Vision: captioningNot measured1297Only Gemma 4 26B A4B measured
LMArena Vision: recognizing things12781191Claude Opus 4.7
LMArena Vision: creative writing13261256Claude Opus 4.7
LMArena Document1495Not measuredOnly Claude Opus 4.7 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 4.7 and Gemma 4 26B A4B
BenchmarkClaude Opus 4.7Gemma 4 26B A4BLeader
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 4.7$0.018Fits in one request
Gemma 4 26B A4B$0.0002Fits in one request

Gemma 4 26B A4B costs less

Repository review

50K input + 3K output tokens

Claude Opus 4.7$0.325Fits in one request
Gemma 4 26B A4B$0.0046Fits in one request

Gemma 4 26B A4B costs less

Agent loop

200K input + 10K output tokens

Claude Opus 4.7$1.25Fits in one request
Gemma 4 26B A4B$0.018Fits in one request

Gemma 4 26B A4B 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 4.7 and Gemma 4 26B A4B
SpecificationAnthropic Claude Opus 4.7Google Gemma 4 26B A4B
MakerAnthropicGoogle
WeightsNo public weights listedOpen weights
ReleasedApril 16, 2026April 3, 2026
Context window1M262K
Input $/Mtok$5.00$0.0765
Output $/Mtok$25.00$0.255
Coding index51.122.8
Independent measurements behind it11
Cost per solved task$2.95Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Claude Opus 4.7 better than Gemma 4 26B A4B for coding?

Pick Claude Opus 4.7 for coding: it scores 51.1 on the Vibe Coding Index against 22.8 for Gemma 4 26B A4B. Pick Gemma 4 26B A4B when output tokens cost 98.0x less, or when the benchmark that matters is IFBench, where it leads.

Which is cheaper, Claude Opus 4.7 or Gemma 4 26B A4B?

Claude Opus 4.7 costs $5.00 per million input tokens and $25.00 per million output tokens; Gemma 4 26B A4B costs $0.0765 and $0.255. Cost per solved task is known for Claude Opus 4.7 only.

Which has the bigger context window, Claude Opus 4.7 or Gemma 4 26B A4B?

Claude Opus 4.7, with 1M tokens against 262K.

How many benchmarks is the Claude Opus 4.7 vs Gemma 4 26B A4B comparison based on?

51 benchmarks have a result for both models as of October 21, 2026. Claude Opus 4.7's coding index rests on 1 independent benchmark and Gemma 4 26B A4B's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next