IBM models: prices and providers
IBM has 9 models on AIModelFind (9 text). 7 have open weights, and 8 providers serve them.
All IBM models
Most searched first. The price is the lowest across providers.
| Model | Lowest price | Providers | Context | License | Released |
|---|---|---|---|---|---|
| Granite 4.2 30B | $0.65/1M | 1 | — | Apache 2.0 | Aug 7, 2026 |
| Granite 3.0 3B A800m Instruct | Free/1M | 1 | — | Apache 2.0 | Oct 3, 2024 |
| Granite 3.0 8B Instruct | $0.25/1M | 2 | — | Apache 2.0 | Oct 2, 2024 |
| Granite 34B Code Instruct | Free/1M | 1 | — | — | — |
| Granite 4.2 3B | $0.12/1M | 1 | — | Apache 2.0 | Aug 7, 2026 |
| Granite 4.0 H Micro | $0.112/1M | 3 | 131K | Apache 2.0 | Sep 16, 2025 |
| Granite-4.0-H-Small | $0.25/1M | 1 | — | Apache 2.0 | Sep 16, 2025 |
| Granite 4.2 8B | $0.25/1M | 6 | 131K | Apache 2.0 | Aug 7, 2026 |
| Granite 8B Code Instruct | Free/1M | 1 | — | — | — |
Where to use IBM models
Providers serving the most IBM models.
IBM at a glance
9 models in total: 9 text. The cheapest is Granite 4.0 H Micro at $0.112 /1M. Granite 4.2 8B has the largest context window, 131K tokens. The most searched is Granite 4.2 30B, with about 2K Google searches a month. The newest is Granite 4.2 30B, released Aug 7, 2026.
Cheapest
- Granite 4.0 H Micro$0.112/1M
- Granite 4.2 3B$0.12/1M
- Granite 3.0 8B Instruct$0.25/1M
- Granite-4.0-H-Small$0.25/1M
- Granite 4.2 8B$0.25/1M
Largest context
Most searched
Newest
- Granite 4.2 30BAug 7, 2026
- Granite 4.2 3BAug 7, 2026
- Granite 4.2 8BAug 7, 2026
- Granite 4.0 H MicroSep 16, 2025
- Granite-4.0-H-SmallSep 16, 2025
IBM models FAQ
We track 9 IBM models: Text 9.
Among IBM's text models, Granite 4.0 H Micro is the cheapest at $0.112 per 1M output tokens.
Yes, 7 IBM models have weights you can download from IBM's Hugging Face page.
8 providers serve IBM models, including Nvidia, Replicate, Hugging Face, AI/ML API, OpenRouter.
Can't find what you're looking for? Contact us