Gemma 4 26B A4B (Reasoning)
GoogleGemmaOpen WeightApache 2.0 · Commercial OK
描述
Gemma 4 26B-A4B is Google DeepMind's Mixture-of-Experts multimodal model with 26 billion total parameters and 3.8 billion activated parameters per token, with a 256K context window. Ranks #6 among open models on Arena AI, outcompeting models 20x its size. Achieves an estimated LMArena text score of 1441. Supports image and text inputs.
發布日期
2026-04-02
參數規模
25.2B
上下文長度
262K
支援模態
image, text, video
能力雷達圖
28
general
25
coding
79
reasoning
53
science估算
90
agents
70
multimodal
Science 在缺少專門科學評測時使用推理能力代理估算。
排行榜排名
基準測試分數 (LLM Stats)
Agents
t2-bench
85.5%自報
Biology
GPQA
82.3%自報
Finance
MMLU-Pro
82.6%自報
General
MMMLU
86.3%自報
LiveCodeBench v6
77.1%自報
MMMU-Pro
73.8%自報
BIG-Bench Extra Hard
64.8%自報
MRCR v2
44.1%自報
Healthcare
MedXpertQA
58.1%自報
Math
AIME 2026
88.3%自報
MathVision
82.4%自報
Humanity's Last Exam
17.2%自報
AA 評測指數
Intelligence Index31.2
Coding Index22.4
Gpqa0.8
Ifbench0.7
Lcr0.6
Tau20.4
Scicode0.4
Hle0.2
Terminalbench Hard0.1
LLM Stats 分類評分
Tool Calling90
Agents90
Biology80
Chemistry80
Finance80
Language80
Legal80
Physics80
General70
Healthcare70
Math70
Multimodal70
Reasoning70
Vision60
Long Context40
定價
輸入價格$0.13 / 1M tokens
輸出價格$0.4 / 1M tokens
混合價格(3:1)$0.198 / 1M tokens
速度
Tokens/秒0.0 tokens/s
首Token延遲0.00s
首回答延遲0.00s
可用提供商
(LS 內部計價單位)| 提供商 | 輸入價格 | 輸出價格 |
|---|---|---|
| Novita | 130K | 400K |