Gemma 4 31B gemma-4-31b-it

google · gemma flagship

#21 overall

params
32.7B
arch
dense
context
256k
license
apache-2.0 open weights
released
Mar 2026
reasoning
yes
downloads/30d
11.5M

# why ranked

Overall: #21 — score 32.9 ● medium — one signal missing, weights renormalized

signalweightinput (0–100)
aa_intelligence 0.50 29.4
arena_elo 0.30
bench_composite 0.20 41.7

missing signals are dropped and the remaining weights renormalized — never imputed.

Coding: provisional — score 48.3 ● low — a single signal; treat with caution

provisional: too few independent signals for a numbered position — the score is shown, but this model sorts after every ranked model.

signalweightinput (0–100)
aa_coding 0.25 48.3
aider_polyglot 0.25
arena_elo_code 0.15
swe_bench_verified 0.35

missing signals are dropped and the remaining weights renormalized — never imputed.

full methodology

# trend

OGM score · last 2 days
OGM score over 2 days
overall rank (up is better)
Overall rank over 2 days

# benchmarks

benchmarkscoresourcedate
AA Coding Index 43.4 Artificial Analysis
AA Intelligence Index 29.4 Artificial Analysis
AIME 2026 0.9 / 1 LLM Stats
Big-bench-extra-hard 0.7 / 1 LLM Stats
Gdpval-aa 783.0 / 3000 LLM Stats

# where to run

providerquantctx$/M in$/M out$/M blendtpsuptime
WandBbf16 256k $0.12$0.35$0.18 100.0%
Venicebf16 250k $0.12$0.36$0.18 99.4%
Chutesfp4 128k $0.12$0.37$0.18 63.4%
DeepInfrafp4 256k $0.12$0.37$0.18 39.2%
DeepInfrafp8 256k $0.13$0.38$0.19 98.7%
SiliconFlowfp8 256k $0.13$0.40$0.20 97.3%
Novitabf16 256k $0.14$0.40$0.21 98.9%
Parasailfp8 256k $0.15$0.40$0.21 95.6%
Phalaunknown 256k $0.15$0.46$0.23 90.4%
ModelRunfp4 256k $0.22$0.55$0.30 100.0%
Togetherunknown 256k $0.28$0.86$0.43 84.2%
Togetherunknown 256k $0.39$0.97$0.53 98.7%
SambaNovaunknown 128k $0.38$1.15$0.57 100.0%

blended = (3·input + output) / 4 per 1M tokens · cheapest first

source aliases
aa
gemma-4-31b
epoch
Gemma 4 31B
openrouter
google/gemma-4-31b-it