ModelBeat
Models/Granite 4.2 8B
All models
I
ibm-granite/granite-4.2-8b

Granite 4.2 8B is IBM’s AI model, released Aug 31, 2026. It costs $0.06/M input and $0.25/M output tokens, and handles a 131K-token context window.

Granite 4.2 8B is a dense reasoning model from IBM. It is suited for mathematics, code generation, multilingual dialogue, and agentic workflows that need multi-step reasoning.

Context
131K
Input
$0.06 / 1M
Output
$0.25 / 1M
Released
Aug 31, 2026
Modalities
texttext

Benchmarks

Scores on standardized evaluations. Higher is better — and rank shows where Granite 4.2 8B lands among all models tracked on Model Beat.

8.0
Intelligence Index
Artificial Analysis
8th percentile of tracked models
8.0
Coding Index
Artificial Analysis
8th percentile of tracked models

Reasoning

2 evals
GPQA Diamond63.1%

Graduate-level scientific reasoning

Humanity's Last Exam9.7%

Frontier of human expert knowledge

Coding

1 evals
SciCode31.5%

Scientific research coding

Changelog

3
  • Granite 4.2 8B: cheapest credible provider via DeepInfra now $0.06 per 1M input tokens (was $0.1, -40%)cheapest provider · Sep 6, 2026
  • Granite 4.2 8B: cheapest credible provider via DeepInfra now $0.25 per 1M output tokens (was $0.15, +67%)cheapest provider · Sep 6, 2026
  • Granite 4.2 8B (IBM) added to the tracker: $0.1/$0.15 per 1M, 131K context.tracked · Aug 31, 2026

Compare Granite 4.2 8B with

Frequently asked questions

What is Granite 4.2 8B?

Granite 4.2 8B is a dense reasoning model from IBM. It is suited for mathematics, code generation, multilingual dialogue, and agentic workflows that need multi-step reasoning.

Who created Granite 4.2 8B?

Granite 4.2 8B is an AI model developed by IBM, released Aug 31, 2026.

How much does Granite 4.2 8B cost?

Granite 4.2 8B is served from $0.06 per million input tokens and $0.25 per million output tokens (cheapest credible provider on OpenRouter: DeepInfra).

What is Granite 4.2 8B's context window?

Granite 4.2 8B supports a context window of up to 131K tokens.

How does Granite 4.2 8B perform on benchmarks?

On standardized evaluations tracked by Epoch AI, Granite 4.2 8B scores 9.7% on Humanity's Last Exam, 31.5% on SciCode, 63.1% on GPQA Diamond.

Benchmark data from Epoch AI (CC BY) and Artificial Analysis; pricing, specs & descriptions from OpenRouter. Data updated Sep 10, 2026.