Granite 4.2 8B

Name: Granite 4.2 8B
Price: 0.082 USD
Author: IBM (Granite)

IBM (Granite) · Granite 4.2 · released 2026-08-07 · data checked 2026-11-01

Granite 4.2 8B is a 8B-parameter dense model from IBM (Granite) (United States), released 2026-08-07 under the Apache-2.0 license. It scores 63.5 on the Small LM Index, #2 of 26 small models. It needs about 6.9 GB of memory at 4-bit with an 8K-token context and costs $0.082 per million tokens through the cheapest API provider.

Small LM Index

63.5

#2 of 26 scored small models

Likely 61.9–65.1 · from 4 of 6 core benchmarks

Parameters

Dense

Price per 1M tokens

$0.082

$0.060 in / $0.15 out · blended 3:1

Context window

131K

tokens

Benchmarks

Each bar shows this model's score. The tick marks the best small model on that benchmark. Sources: 8 reported by IBM (Granite). How we label sources.

MMLU-Pro Lab-reported Knowledge74.0

Class best: Qwen3.5 9B 82.5

GPQA Diamond Lab-reported Knowledge64.1

Class best: Qwen3.5 9B 81.7

AIME Lab-reported Math · AIME 202586.7

Class best: Apriel 1.6 15B Thinker 88.0

LiveCodeBench Lab-reported Coding · LiveCodeBench v673.2

Terminal-Bench Lab-reported Agentic · Terminal-Bench 2.120.6

IFBench Lab-reported Instruction following · Prompt-level79.3

BFCL Lab-reported Agentic · BFCL v452.4

Class best: Hunyuan 7B Instruct 70.8

HMMT Lab-reported Math · HMMT Feb 202578.3

Not reported: estimated

What the index model expects from this model's other scores. Not a measurement.

Humanity's Last Exam≈ 13

SWE-bench Verified≈ 52

Specifications

Lab: IBM (Granite) (US · United States)
Architecture: Dense transformer
Total parameters: 8B
Active parameters: All (dense)
Size class: Small
Context window: 131K tokens
Input modalities: text
Reasoning: Instruct (no thinking)
License: Apache-2.0
Openness: Open weights
Memory to run: ~6.9 GB at 4-bit, 8K context (more sizes)
Hugging Face: 71.2K downloads (30 days) · 112 likes
API providers: 2 on OpenRouter
API price: $0.060 in / $0.15 out per 1M (OpenRouter, 2026-11-03)

Weights on Hugging Face ↗Try on OpenRouter ↗

Compare with

Granite 4.2 8B vs Ministral 3 8B (8B, index 55.2)
Granite 4.2 8B vs Gemma 4 E4B (8B, index 38.2)
Granite 4.2 8B vs Llama 3.1 8B (8B, index 21.3)
Granite 4.2 8B vs Ling 3.0 Tiny (7.9B, index 51.9)
Granite 4.2 8B vs DeepSeek-R1-0528-Qwen3-8B (8.2B, index 53.1)
Granite 4.2 8B vs Qwen3 8B (8.2B, index 48.5)

Ranked in

#5 in Best LLMs that run on an 8 GB GPU or laptop
#6 in Best LLMs that run on a 16 GB GPU
#9 in Best cheap LLM APIs (open models)
#9 in Best American open models under 100B

Sources

Something wrong here? Report it.

Run it locally

Uses ibm-granite/granite-4.2-8b-GGUF (Q4_K_M). Swap the tag for another size: BF16, Q2_K, Q3_K_L, Q3_K_M, Q3_K_S, Q4_0, Q4_1, Q4_K_M….

Ollama

ollama run hf.co/ibm-granite/granite-4.2-8b-GGUF:Q4_K_M

llama.cpp

llama-server -hf ibm-granite/granite-4.2-8b-GGUF:Q4_K_M

Then open http://localhost:8080 for a chat window.

MLX (Mac)

pip install -U mlx-lm
mlx_lm.chat --model lmstudio-community/granite-4.2-8b-MLX-4bit

vLLM (GPU server)

vllm serve ibm-granite/granite-4.2-8b

Full-precision weights, for a server GPU.

Open in LM Studio

Memory needed: weights, context cache and runtime buffers.

Quantization	4K context	32K context	131K context
16-bit F16	17 GB	22 GB	38 GB
8-bit Q8_0	9.8 GB	15 GB	31 GB
4-bit Q4_K_M	6.2 GB	11 GB	27 GB

Check it against your GPU, Mac or phone →