IBM Granite models: open weights at near the lowest price here
IBM
Services found: 2
| # | AI model | Capabilities | API $/1M (in / out) | Context window | Price / mo | Output $/1M | Overview | Visit site |
|---|---|---|---|---|---|---|---|---|
| 1 |  Granite 4.0 Micro IBM |    | $0.017 / $0.112 | 131K tokens | API only | 0.112 | View | |
| 2 | Granite 4.2 8B IBM | | $0.06 / $0.25 | 131.1K tokens | API only | 0.25 | View |
Browse by need
IBM Granite family is deliberately small and cheap: Granite 4.0 Micro at $0.017 input and $0.112 output per 1M tokens, and Granite 4.2 8B a step above. Both are open weight.
Granite is positioned for enterprises that need auditable provenance and on-premise deployment more than they need frontier reasoning — regulated industries where the model must run inside a controlled boundary and its training story must survive review.
Cheapest input rate$0.017 per 1M, the lowest in this rating.
Governance storyBuilt for environments where provenance is audited.
Open weightsSelf-hosting is the expected deployment.
Models from this vendor
2 IBM models in this rating. Prices are input and output per 1M tokens from the vendor official price list; for dirhams multiply by 3.6725. Tap a card to open the model page.
How to choose
- Benchmark against a comparable model from a larger vendor before committing production traffic — line-ups this size get less independent testing.
- Start with Granite 4.0 Micro at $0.017 / $0.112 per 1M tokens; it is the cheapest entry point here.
- Move up to Granite 4.2 8B at $0.06 / $0.25 only where the cheaper variant visibly fails, not by default.
- Check the context window against your longest input — the largest here is Granite 4.2 8B at 131.1K tokens.
Common mistakes
Assuming parity with the majorsOn the hardest tasks the frontier labs remain ahead; compare on your own cases.
Skipping the exit planA smaller vendor can change terms or availability faster. Identify a fallback model before you depend on this one.
Who this suits
Cost-driven teamsGranite 4.0 Micro at $0.017 / $0.112 undercuts most mainstream options.
EvaluatorsA useful comparison point when benchmarking the models against larger line-ups.
Regional usersRelevant where this vendor platform or language coverage matters more than raw capability.
Related filters
Text & chat228 models for text generation and chat
Reasoning169 models that think before answering
Tool calling166 models that call your functions
Low latency73 small, fast models for volume
Alibaba (Qwen)34 models in this rating
OpenAI30 models in this rating
Google DeepMind17 models in this rating
Frequently asked questions
What is Granite for?
Regulated enterprise workloads where deployment control outweighs raw capability.
Is Granite competitive with frontier models?
No, and it is not trying to be. It targets small, controllable and cheap.
Can I run it on-premise?
Yes — that is the primary use case.
How many IBM models are in this rating?
2, covering text & chat, reasoning, tool calling.
Is IBM a good default choice?
For most teams no — start with a mainstream vendor and use this one where its pricing or specialisation gives a concrete advantage.
Frequently searched
By capabilityReasoningCodingVisionLong context 1M+Free tierOpen weightsLow latencyRealtime voiceImage generation
By vendorAnthropic (13)OpenAI (30)Google DeepMind (17)xAI (8)Meta (6)DeepSeek (9)Mistral AI (11)Alibaba (Qwen) (34)Z.ai (12)Moonshot AI (5)
Rate this page
Help us improve