Skip to main content

Granite 4.0 Micro

IBMOpen WeightApache 2.0 · Commercial OK

Description

A preliminary version of the smallest model in the upcoming Granite 4.0 family, released May 2025. It utilizes a novel hybrid Mamba-2/Transformer, fine-grained mixture of experts (MoE) architecture (7B total parameters, 1B active at inference). This preview version is partially trained (2.5T tokens) but demonstrates significant memory efficiency and performance potential, validated for at least 128K context length without positional encoding.

Release Date
2025-09-22
Parameters
7.0B
Context Length
—
Modalities
—

Capability Radar

60
general
80
coding
60
reasoning
51
scienceest.
66
agents
0
multimodal

Science is estimated from LLM Stats science scores or reasoning when no dedicated science benchmarks are available.

Rankings

No ranking data available

Benchmark Scores (LLM Stats)

(LLM Stats (zeroeval))

Chat

IFEvalGoogle Research (2023)63.0%SR

General

MMLU60.4%SR
TruthfulQA58.1%SR
AlpacaEval 2.035.2%SR
Arena Hard26.7%SR
PopQA22.9%SR

Math

GSM8k70.1%SR

Reasoning

HumanEvalOpenAI (2021)82.4%SR
HumanEval+78.3%SR
BIG-Bench Hard55.7%SR
DROP46.2%SR

Safety

AttaQ86.1%SR

AA Evaluation Indices

(Artificial Analysis)

No AA evaluation data available

LLM Stats Category Scores

(LLM Stats (zeroeval))
Safety
90
Code
80
Instruction Following
60
Language
60
Legal
60
Math
60
Structured Output
60
Finance
60
General
60
Healthcare
60
Reasoning
50
Chat
40
Creativity
30
Writing
30

Pricing

No pricing data available

Speed

No speed data available

Provider Price Ranking

No provider data available

External Sources