Granite 4.2 8B
by IBMtext->text2 endpoints
Neura Intelligence Index
37.1/ 100
Rank
#241—
Confidence
mediumBenchmarks
5(13% coverage)
Overview
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. It supports full, low-effort, and non-thinking modes.
The model supports 12 languages: English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, and Chinese.
Capabilities
Text generation
Tool use / Function calling
Structured output
Extended reasoning
Prompt caching
Long context
Top-P sampling
Stop sequences
Deterministic seed
Log probabilities
Modalities
Input
Text
Output
Text
Technical Specifications
Context Window
131.1K tokens
Max Output
118.0K tokens
Tokenizer
Other
Uptime (24h)
9999.4%
Parameters
8000000000B
License
apache_2_0Open Weights
Country
🇺🇸US
Release Date
August 25, 20263w
Supported Parameters
frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningreasoning_effortrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p
Benchmark Performance
coding
| Benchmark | Score | Source |
|---|---|---|
| SWE-bench Verified | 47.7 | verified |
| SciCode | 36.1 | verified |
| SWE-Bench Pro | 19.1 | verified |
math
| Benchmark | Score | Source |
|---|---|---|
| AIME 2025 | 86.7 | verified |
reasoning
| Benchmark | Score | Source |
|---|---|---|
| GPQA Diamond | 64.1 | verified |
Performance Over Time
Pricing
Per 1M tokens
Input$0.06
Output$0.25
Cache read$0.01
Added
August 31, 2026
Last synced 9/18/2026