Granite 4.0 Micro
Summary
Granite 4.0 Micro is an open-weight language model from IBM, released on 20 Oct 2025. It accepts text and generates text, with a 131K-token context window and up to 118K output tokens per response.
Granite-4.0-H-Micro is a 3B parameter from the Granite 4 family of models. These models are the latest in a series of models released by IBM. They are fine-tuned for long...
Last verified 19 Nov 2026. Every figure on this page links to its source.
Specifications
| Developer | IBM |
|---|---|
| Release date | 20 Oct 2025 [OpenRouter] |
| Status | Generally available |
| Weights | Open weights [Hugging Face] |
| License | apache-2.0 [Hugging Face] |
| Parameters | 3.2B [Hugging Face] |
| Context window | 131K (131,000 tokens) [OpenRouter] |
| Max output | 118K tokens [OpenRouter] |
| Input | Text [OpenRouter] |
| Output | Text [OpenRouter] |
| Reasoning mode | No [OpenRouter] |
| Tool calling | No [OpenRouter] |
| Structured output | No [OpenRouter] |
| Hugging Face | ibm-granite/granite-4.0-h-micro [Hugging Face] |
Granite 4.0 Micro API pricing by provider
USD per million tokens. Sorted by input price.
We do not currently track a paid API listing for this model.
Running Granite 4.0 Micro locally
Estimated memory, from the parameter count and the architecture in config.json on Hugging Face. Includes about 0.8 GB of runtime overhead.
| Quantization | Weights | Total, 8K context | Total, 32K context |
|---|---|---|---|
| Q4_K_M (4-bit) | 1.9 GB | 2.8 GB | 3.0 GB |
| Q8_0 (8-bit) | 3.4 GB | 4.3 GB | 4.5 GB |
| FP16 / BF16 (16-bit) | 6.4 GB | 7.2 GB | 7.5 GB |
At 4-bit with an 8K conversation, Granite 4.0 Micro needs about 2.8 GB, so it fits on a graphics card with 8 GB or more, such as the NVIDIA GeForce RTX 4060 8 GB. Check your own hardware, or see Running local LLMs on an 8 GB GPU on our sister site AIPCs.
Frequently asked questions
What is the context window of Granite 4.0 Micro?
Granite 4.0 Micro has a context window of 131,000 tokens (131K) and can generate up to 117,900 tokens in a single response, according to OpenRouter.
Is Granite 4.0 Micro open source?
Granite 4.0 Micro is an open-weight model: its weights are published on Hugging Face as ibm-granite/granite-4.0-h-micro under the apache-2.0 license. Check the license terms before commercial use.
When was Granite 4.0 Micro released?
IBM released Granite 4.0 Micro on 20 Oct 2025, according to OpenRouter.
How much VRAM does Granite 4.0 Micro need?
About 2.8 GB at 4-bit (Q4_K_M) with an 8K-token conversation, and about 4.3 GB at 8-bit. The 4-bit model fits on a graphics card with 8 GB or more, such as the NVIDIA GeForce RTX 4060 8 GB. These are estimates from the model's published parameter count and architecture.
Similar models from other labs
Comparable on output type, weights, release date, price and context window. Prices are the lowest listed, USD per million tokens.
| Model | Lab | Released | Context | Input $/1M | Output $/1M | Weights |
|---|---|---|---|---|---|---|
| Qwen3 VL 32B Instruct | Alibaba (Qwen) | 23 Oct 2025 | 131K | – | – | Open |
| Qwen3 VL 8B Thinking | Alibaba (Qwen) | 14 Oct 2025 | 131K | – | – | Open |
| Nvidia Nemotron Nano 12B V2 VL | NVIDIA | 28 Oct 2025 | 131K | $0.20 | $0.60 | Open |
| GPT OSS Safeguard 20B | OpenAI | 29 Oct 2025 | 131K | $0.07 | $0.20 | Open |
| gpt-oss-safeguard-120b | OpenAI | 29 Oct 2025 | 131K | $0.15 | $0.60 | Open |
| llama-3.1-nemotron-safety-guard-8b-v3 | NVIDIA | 28 Oct 2025 | 128K | – | – | Open |
Change history
- No longer listed on openrouter