Skip to main content
Meo

Google: Gemma 4 26B A4B

Google: Gemma 4 26B A4B is google's model. On the un-leaked meo benchmark it ranks #54 of 82 with 55.2% accuracy and an Effective Value (𝕍) of 0.0001 β€” earning a correct answer for about $0.0043 in 258s each.

as of 2026-10-18

meo rank
#54
of 82
Accuracy
55.2%
overall
𝕍 index (top=100)
0.003
raw 𝕍 0.0001 Β· N=10
$ / correct
$0.0043
258s each

Per-domain accuracy

139 correct of 250 questions asked across 11 domains Β· answer rate 100% Β· macro accuracy 57.4%. Headline accuracy is micro-averaged per meo’s methodology; per-domain question counts (correct/asked) are shown so you can judge significance.

Theory of mind
100%3/3
Regex automaton
81%13/16
Base-rate bias
80%33/41
Tape machine
80%12/15
Watson-Glaser
75%3/4
CSP (unsatisfiable)
69%11/16
Perceptual illusions
47%48/102
Logic / math / CS
44%15/34
Framework bias
33%1/3
State tracking
0%0/1
Long arithmetic
0%0/15

Efficiency

Cost per correct
$0.0043
Seconds per correct
258
Tokens per correct
11,369
Correct answers
139 / 250
Total benchmark cost
$0.60
Avg tokens / item
6,321

Model details

Released
2026-04-03
Weights
Open
Context
262K
Max output
236K
Knowledge cutoff
β€”
Price in
$0.08/M
Price out
$0.26/M
Modality
text+image+video->text
View on OpenRouter β†—Hugging Face β†—

Sources: openrouter Β· meo

FAQ

How accurate is Google: Gemma 4 26B A4B ?

Google: Gemma 4 26B A4B scores 55.2% overall on the meo benchmark (rank #54 of 82), an un-leaked private holdout graded objective-first.

What is Google: Gemma 4 26B A4B 's Effective Value?

Its Effective Value (𝕍) is 0.0001 at N=10, Ο‰=1, Ξ΄=1.5 β€” a fused metric of accuracy, speed, cost, and error-cascade. 𝕍 rankings shift with task depth.

How much does Google: Gemma 4 26B A4B cost per correct answer?

About $0.0043 per correct answer (11,369 tokens, 258 seconds each) at benchmark-time pricing via OpenRouter.

Is Google: Gemma 4 26B A4B open weights?

Yes β€” Google: Gemma 4 26B A4B is released with open weights.

Benchmark updates affecting this model

  • 2026-06-09majorlogic_math_cs ground truth was broken β€” culled and rebuilt with solver-verified items

    A 5-lab frontier jury revealed that 29 of 37 LLM-authored logic_math_cs items had broken ground truth: on most, five independent frontier models unanimously concluded the clue sets were contradictory while the stored answer claimed a number, and on eight, all five agreed on the SAME different number (author arithmetic errors). Models were being marked wrong for being right. The 29 were retired (full audit published in the repo) and the domain rebuilt with generator-as-oracle items: an embedded constraint solver PROVES each puzzle has exactly one solution (or provably none), verified by an independent exhaustive enumerator. A 3-round difficulty study found frontier models saturate consistent human-scale logic puzzles, so items use anonymous global constraints and carry a measured difficulty gradient.

  • 2026-06-08majorCorrected base_bias grading (LLM-equivalence judge fix)

    The LLM-equivalence grader uses a reasoning model whose verdict token budget was too small, so its YES/NO answer was emitted as empty and silently counted as a failure. This understated the base_bias domain to ~7% for every model when the true accuracy is ~85-90%. We raised the judge budget, excluded its reasoning tokens from the verdict, switched to verbosity-robust conclusion matching, and re-graded the whole board.

Full benchmark changelog β†’

Embed this ranking

Show Google: Gemma 4 26B A4B ’s meo rank on your site β€” the badge links back here.

Related: all Google models Β· Theory of mind domain Β· best Effective Value Β· biggest movers Β· methodology Β· full leaderboard.