Back to Explore
Google: Gemma 4 26B A4B
Google·Released 2026-04-03
Open Source26B262K ctxApache 2.0Multimodal
About
Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Despite 25.2B total parameters, only 3.8B activate per token during inference — delivering near-31B quality at...
Related Models
Pricing
Input
$0.06
per 1M tokens
Output
$0.33
per 1M tokens
Blended
$0.13
per 1M tokens
Cheaper than 82% of models. Median price is $0.70/1M tokens.
Cost Calculator
Tokens per day1M
100K100M
Daily
$0.13
Monthly
$3.83
vs. Similar Models
Step 3.7 FlashQ:0.0
$0.44+244%
DeepSeek V3.1 Terminus (Reasoning)Q:+0.1
$1.91+1401%
o3Q:+0.1
$3.50+2645%
GPT-5.4 nano (medium)Q:-0.1
$0.46+263%
Performance
Context Window
262K
tokens
Larger than 60% of models
Max Output
262K
tokens
100% of context
Context Window Comparison
Kimi K2.6
262KSame
Kimi K2.7 Code
262KSame
Nex-N2-Pro
262KSame
Benchmarks
MMLU-ProNot evaluated
GPQA Diamond
79.2%
HLE
18.3%
LiveCodeBenchNot evaluated
SciCode
40.0%
TerminalBench Hard
8.3%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025Not evaluated
IFBench
72.4%
Long Context Recall
55.7%
Tau2
43.6%
Market AverageTop Score
Open Source
apache-2.026BGGUF / GPTQ / AWQ
Downloads
1.0M
Likes
338
VRAM (FP16)
24-48 GB
GPU
A6000 / M3 Ultra