Saltar al contenido principal
AITier

Granite 4.2 8B

IBMOpen WeightApache 2.0 · Uso Comercial

Descripción

IBM Granite 4.2 8B is a mid-size dense decoder-only reasoning LLM in the Granite 4.2 family. It supports native think, non-think, and low-effort modes, tool calling, and a 128K native context window with extension to 512K. Like the 30B, it received additional agentic reinforcement learning for multi-step tool use. Open weights on Hugging Face under Apache 2.0. Text-only (not multimodal).

Fecha de lanzamiento
2026-08-25
Parámetros
8.0B
Longitud del contexto
131K
Modalidades
text

Radar de capacidades

11
general
24
coding
63
reasoning
41
science
40
agents
0
multimodal

Rankings

Dominio#PosiciónPuntuaciónFuente
Capacidad agéntica102
38.0
LS
Ranking de codificación295
35.0
AA
Ranking general528
15.0
AA
Ciencia298
39.0
AA

Puntuaciones de benchmarks (LLM Stats)

(LLM Stats (zeroeval))

Agents

TAU3-Bench58.1%Aut.

General

Arena-Hard v265.2%Aut.

Instruction Following

IFBench79.3%Aut.

Language

MMLU-Pro74.0%Aut.

Long Context

RULER 64k81.0%Aut.
RULER 128k71.4%Aut.

Math

AIME 202586.7%Aut.
HMMT2578.3%Aut.

Reasoning

LiveCodeBench v673.2%Aut.
GPQANYU + Cohere + Anthropic (2023)64.1%Aut.
SWE-Bench Verified47.7%Aut.
ProfBench41.2%Aut.
Bird-SQL (dev)41.1%Aut.
SciCode36.1%Aut.
SWE-bench Multilingual30.8%Aut.
Terminal-Bench 2.120.6%Aut.
SWE-Bench ProPrinceton NLP (2024)19.1%Aut.

Tool Calling

BFCL-V452.4%Aut.

Índices de evaluación AA

(Artificial Analysis)
Gpqa(NYU + Cohere + Anthropic (2023))
63.1
Lcr(Artificial Analysis)
45.0
Scicode(UIUC + Argonne National Lab (2024))
31.5
Coding Index(Artificial Analysis)
22.4
Terminalbench V2 1
18.4
Intelligence Index(Artificial Analysis)
11.1
Hle(Center for AI Safety + Scale AI (2025))
9.7
Tau Banking
7.6
Terminalbench V4 0
0.0

Puntuaciones por categoría LLM Stats

(LLM Stats (zeroeval))
Instruction Following
80
Long Context
80
Language
70
Legal
70
Math
70
Finance
70
Healthcare
70
Creativity
70
Writing
70
General
60
Physics
50
Reasoning
50
Frontend Development
50
Biology
50
Chemistry
50
Knowledge
40
Agents
40
Tool Calling
40
Code
30

Precios

Precio de entrada$0.06 / 1M tokens
Precio de salida$0.25 / 1M tokens
Precio mixto (3:1)$0.107 / 1M tokens
Precio de lectura caché$0.015 / 1M tokens

Velocidad

Tokens/seg97.6
Retraso del primer token0.33s
Tiempo hasta la respuesta20.82s

Ranking de Precios por Proveedor

Ranking de Precios por Proveedor

6 proveedores

Más barato: DeepInfraMás caro: CoreWeave
ProveedorEntradaSalida
1DeepInfraMás barato
$0
$0
2IBMPRINCIPAL
$0.06
$0.25
3OpenRouter
$0.06
$0.25
4Kilo Gateway
$0.06
$0.25
5NanoGPT
$0.1
$0.15
6CoreWeave
$0.1
$0.15

Comparar precios entre diferentes proveedores de API para este modelo.

Fuentes externas

LLM StatsArtificial Analysis