IBM models

IBM develops the Granite family of open models for enterprise use, covering language, vision, speech, guardrails, embeddings and time-series forecasting. Granite 4.2 (August 2026) offers dense 3B, 8B and 30B language models with built-in reasoning modes. Models are available on Hugging Face, Ollama and IBM watsonx.ai.

2 models trackedData checked October 10, 2026

Lineup

Model	Status	Weights	Tier	Context	Input / 1M	Output / 1M	Released
Granite 4.2 30BGranite 4.2	Current	Open weights	Efficient	128K	$0.16	$0.65	Aug 25, 2026
Granite 4.2 8BGranite 4.2	Current	Open weights	Compact	128K	$0.06	$0.25	Aug 25, 2026

The lowest-priced current model is Granite 4.2 8B at $0.06 input and $0.25 output per million tokens. The most capable tier in the lineup is efficient, led by Granite 4.2 30B. The longest context window is 128K tokens (Granite 4.2 30B).

Frequently asked questions

What models does IBM offer?

As of October 10, 2026, we track 2 IBM models: Granite 4.2 30B, Granite 4.2 8B. Status, prices and context windows for each are in the table on this page.

How do I access IBM models?

Through IBM's developer platform (huggingface.co/ibm-granite), and for many models through cloud marketplaces and inference providers listed on each model page.

Does IBM release open-weight models?

Granite language models are released under Apache 2.0.

Other providers

OpenAI Anthropic Google xAI Amazon (AWS)Meta Perplexity Mistral AI DeepSeek Alibaba (Qwen)Microsoft (Phi)NVIDIA Moonshot AI Z.ai (Zhipu AI)MiniMax Ai2 (Allen Institute for AI)

Still choosing a model?

Send us your use case and constraints. We'll come back with a shortlist and an evaluation plan you can run on your own data.

Free and vendor-neutral: no model provider pays for placement.
Your details go only to partners you agree to.