modelgrep
I

IBM models

2models tracked · ranked by intelligence, speed & price

Quick answer · Updated June 2026

IBM's smartest model is Granite 4.1 8B (12.4 on the Intelligence Index), its fastest is Granite 4.1 8B at 94 tokens/sec, and its cheapest is Granite 4.0 Micro at $0.017 per million input tokens. All 2 IBM models are compared below by intelligence, speed, latency, context and price.

Smartest
12.4
ibm-granite/granite-4.1-8b
Fastest
94 t/s
ibm-granite/granite-4.1-8b
Cheapest
$0.017/M
ibm-granite/granite-4.0-h-micro
ModelIntelIn $/MContext
granite-4.1-8b
ToolsJSON
12.4$0.050131K
granite-4.0-h-micro
7.7$0.017131K