Granite-4.0-micro (IBM)

Ultra-small language model built for efficiency, safety, and enterprise-friendly use.

• Micro by design. A very small model from IBM’s Granite 4.0 family, optimized for lightweight NLP tasks where cost, latency, and predictability matter more than deep reasoning.

• Task-focused. Best suited for classification, extraction, routing, tagging, and short-form text generation rather than open-ended chat.

• Fast and cheap to run. Minimal memory footprint makes it practical on CPUs, edge devices, and high-throughput systems.

• Enterprise-safe defaults. Designed with conservative behavior, stable outputs, and clear constraints—well aligned with production and regulated environments.

• Straightforward integration. Single model, standard inputs, no complex tooling required.

Why pick it for Norman AI?

Granite-4.0-micro is a good fit when you need reliable text processing at scale with minimal compute. Ideal for pipelines, automation, and background tasks where speed and cost dominate.

messages = [
    {"role": "system", "content": "You are a helpful AI assistant."},
    {"role": "user", "content": "Can you provide ways to eat combinations of bananas and dragonfruits?"},
    {"role": "assistant",
     "content": "Sure! Here are some ways to eat bananas and dragonfruits together"},
    {"role": "user", "content": "What about solving an 2x + 3 = 7 equation?"},
]

response = await norman.invoke(
    {
        "model_name": "granite-4.0-micro",
        "inputs": [
            {
                "display_title": "Prompt",
                "data": messages
            }
        ]
    }
)

messages = [
    {"role": "system", "content": "You are a helpful AI assistant."},
    {"role": "user", "content": "Can you provide ways to eat combinations of bananas and dragonfruits?"},
    {"role": "assistant",
     "content": "Sure! Here are some ways to eat bananas and dragonfruits together"},
    {"role": "user", "content": "What about solving an 2x + 3 = 7 equation?"},
]

response = await norman.invoke(
    {
        "model_name": "granite-4.0-micro",
        "inputs": [
            {
                "display_title": "Prompt",
                "data": messages
            }
        ]
    }
)

messages = [
    {"role": "system", "content": "You are a helpful AI assistant."},
    {"role": "user", "content": "Can you provide ways to eat combinations of bananas and dragonfruits?"},
    {"role": "assistant",
     "content": "Sure! Here are some ways to eat bananas and dragonfruits together"},
    {"role": "user", "content": "What about solving an 2x + 3 = 7 equation?"},
]

response = await norman.invoke(
    {
        "model_name": "granite-4.0-micro",
        "inputs": [
            {
                "display_title": "Prompt",
                "data": messages
            }
        ]
    }
)

messages = [
    {"role": "system", "content": "You are a helpful AI assistant."},
    {"role": "user", "content": "Can you provide ways to eat combinations of bananas and dragonfruits?"},
    {"role": "assistant",
     "content": "Sure! Here are some ways to eat bananas and dragonfruits together"},
    {"role": "user", "content": "What about solving an 2x + 3 = 7 equation?"},
]

response = await norman.invoke(
    {
        "model_name": "granite-4.0-micro",
        "inputs": [
            {
                "display_title": "Prompt",
                "data": messages
            }
        ]
    }
)