NANI-Nithin/granite-4.2-3b-GGUF

Granite-4.2-3B GGUF (llama.cpp)

Quantized GGUF files for Granite-4.2-3B, optimized for deployment with llama.cpp. These are the exact quantized versions of the Granite-4.2-3B model, packaged as GGUF format for efficient inference.

Model Overview

Granite-4.2-3B is IBM's compact reasoning model in the Granite 4.2 family. Despite its small parameter count of 3B, it delivers strong performance on reasoning-intensive tasks by leveraging built-in <think>...</think> chain-of-thought reasoning. It supports flexible thinking modes — full thinking (default), non-thinking, and low-effort — allowing users to balance depth vs. latency on a per-query basis.

Key Capabilities

• Built-in Reasoning: Native chain-of-thought that significantly improves performance on math, coding, and complex multi-step problems
• Flexible Thinking Modes: Seamlessly switch between full thinking, non-thinking, and low-effort modes within a single model
• Reasoning-Augmented Tool Calling: The model reasons about which tools to invoke and why, producing more accurate function calls
• 512K Context Window: Supports long documents, multi-turn conversations, and complex agentic workflows
• Apache 2.0 Licensed: Fully open for commercial and research use

Available Quantizations

File	Size	Quality	Recommended Use
granite-4.2-3b-BF16.gguf	6.82 GB	Full precision source. Every quant below is cut from this file.	Original model, no quantization
granite-4.2-3b-Q8_0.gguf	3.63 GB	Effectively lossless. Use when disk and RAM are not the constraint.	Highest quality, less compression
granite-4.2-3b-Q6_K.gguf	2.80 GB	Near-lossless; the last stop before quality becomes measurable.	Balanced quality/size
granite-4.2-3b-Q5_K_M.gguf	2.43 GB	Very good quality, noticeably smaller than Q6_K.	Good trade-off
granite-4.2-3b-Q5_K_S.gguf	2.38 GB	Slightly smaller than Q5_K_M for a slight quality cost.	Smaller footprint
granite-4.2-3b-Q5_1.gguf	2.58 GB	Legacy. Prefer Q5_K_M.	Historical compatibility
granite-4.2-3b-Q5_0.gguf	2.38 GB	Legacy. Prefer Q5_K_M.	Historical compatibility
granite-4.2-3b-Q4_K_M.gguf	2.09 GB	The usual default. Best quality-per-byte for most people.	Default choice
granite-4.2-3b-Q4_K_S.gguf	2.00 GB	A little smaller than Q4_K_M, a little worse.	Smaller footprint
granite-4.2-3b-IQ4_NL.gguf	1.99 GB	Non-linear 4-bit; good on hardware without fast K-quant kernels.	Specialized hardware
granite-4.2-3b-IQ4_XS.gguf	1.89 GB	Best sub-4.5bpw option; usually beats Q4_K_S at a smaller size.	Size-critical
granite-4.2-3b-Q4_1.gguf	2.18 GB	Legacy. Prefer Q4_K_M.	Historical compatibility
granite-4.2-3b-Q4_0.gguf	1.98 GB	Legacy round-to-nearest. Prefer Q4_K_M unless a runtime needs this.	Historical compatibility
granite-4.2-3b-Q3_K_L.gguf	1.84 GB	Small, with real quality loss. Usable when RAM is tight.	Tight RAM constraints
granite-4.2-3b-Q3_K_M.gguf	1.71 GB	Smaller again; noticeable degradation.	Memory-constrained
granite-4.2-3b-IQ3_M.gguf	1.61 GB	Strong at ~3.7bpw, clearly better than Q3_K_M.	Quality-conscious sizing
granite-4.2-3b-IQ3_S.gguf	1.57 GB	Slightly smaller than IQ3_M.	Compact version
granite-4.2-3b-Q3_K_S.gguf	1.56 GB	Aggressive. Prefer IQ3_M at a similar size.	Maximum compression
granite-4.2-3b-IQ3_XS.gguf	1.50 GB	Aggressive but coherent.	Extreme compression
granite-4.2-3b-IQ3_XXS.gguf	1.39 GB	Very aggressive; the last coherent step down.	Extreme compression
granite-4.2-3b-Q2_K.gguf	1.36 GB	Very small, heavily degraded. For experimentation.	Experimental only
granite-4.2-3b-IQ2_M.gguf	1.26 GB	The smallest size most people find usable.	Memory-constrained
granite-4.2-3b-Q2_K_S.gguf	1.27 GB	Smaller than Q2_K, at a further quality cost.	Even smaller
granite-4.2-3b-IQ2_S.gguf	1.18 GB	Below the usual usability line.	Extreme compression
granite-4.2-3b-IQ2_XS.gguf	1.11 GB	Experimental.	Experimental only
granite-4.2-3b-IQ2_XXS.gguf	1.02 GB	Experimental.	Experimental only
granite-4.2-3b-Q2_0.gguf	1.16 GB	Extreme, group-64. Included for completeness.	Historical compatibility
granite-4.2-3b-IQ1_M.gguf	0.92 GB	Extreme. Expect substantial degradation.	Maximum compression
granite-4.2-3b-IQ1_S.gguf	0.87 GB	Extreme. Expect substantial degradation.	Maximum compression
granite-4.2-3b-Q1_0.gguf	0.65 GB	Extreme. Included for completeness.	Historical compatibility

All files are cut from the BF16 source (6.82 GB). Each quant is independently uploaded and deleted immediately after successful upload to minimize peak disk usage.

Model Architecture

Granite-4.2-3B is built on a decoder-only dense transformer architecture:

• Attention: Grouped Query Attention (GQA) with 40 attention heads and 8 KV heads
• Position Embedding: Rotary Position Embedding (RoPE) with θ = 10,000,000
• Feed-Forward: MLP with SwiGLU activation (hidden size 8192)
• Normalization: RMSNorm (ε = 1e-5)
• Embeddings: Separate input/output embeddings (not tied)
• Precision: bfloat16 (source)

Inference

Generation Parameters

Important: Use temperature=1.0 and top_p=0.95 across all tasks and serving backends, including general chat, reasoning, and tool calling.

Parameter	Value	Notes
temperature	1.0	Required for all modes
top_p	0.95	Nucleus sampling threshold
max_new_tokens	8192	Thinking mode (increase for complex reasoning)
max_new_tokens	2048	Non-thinking mode
do_sample	True	Required when temperature > 0

Thinking Modes

Mode	Template Parameters	Behavior
Thinking (default)	enable_thinking=True	Full chain-of-thought reasoning inside <think>...</think>
Non-thinking	enable_thinking=False	Direct answer with no reasoning overhead
Low-effort	enable_thinking=True, low_effort=True	Brief reasoning for simpler queries

Serving with llama.cpp

Basic Usage

# Download a quantization (recommended: Q4_K_M)
huggingface-cli download NANI-Nithin/granite-4.2-3b-GGUF granite-4.2-3b-Q4_K_M.gguf --local-dir .

# Run inference
llama-cli -m granite-4.2-3b-Q4_K_M.gguf -p "Hello, how are you?"

Or use the Hugging Face Hub integration:

# Serve with llama-server
llama-server -hf NANI-Nithin/granite-4.2-3b-GGUF:Q4_K_M

# Run with model path
llama-cli -hf NANI-Nithin/granite-4.2-3b-GGUF:Q4_K_M -p "Hello"

API Usage

# Start the server
llama-server -hf NANI-Nithin/granite-4.2-3b-GGUF:Q4_K_M

# Query with curl
curl -X POST http://localhost:8080/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "granite-4.2-3b-Q4_K_M",
    "messages": [{"role": "user", "content": "Explain quantum computing in simple terms"}],
    "temperature": 1.0,
    "top_p": 0.95,
    "max_tokens": 8192
  }'

Quick Start Example

# Download the model
huggingface-cli download NANI-Nithin/granite-4.2-3b-GGUF granite-4.2-3b-Q4_K_M.gguf

# Run inference
./llama-cli -m granite-4.2-3b-Q4_K_M.gguf -p "What is the Riemann hypothesis?"

Technical Details

• Source Model: ibm-granite/granite-4.2-3b
• Author: Granite Team, IBM
• License: Apache 2.0
• Languages: English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, Chinese
• Architecture: GraniteForCausalLM (decoder-only dense transformer)
• Context Length: 131K tokens (natively supports 128K, extended to 512K)
• Parameters: 3B
• Created: August 25, 2026

Usage Notes

1. Disk Space: Download one quantization at a time. Each file ranges from 0.65 GB (Q1_0) to 6.82 GB (BF16 source).
2. Memory Requirements: Varies by quantization. Q4_K_M requires ~2 GB VRAM, Q8_0 requires ~4 GB VRAM.
3. Performance: All files are optimized for llama.cpp with appropriate kernel selections for each quantization.
4. Thinking Mode: Enable enable_thinking=True to get chain-of-thought reasoning. Set to False for faster, direct answers.

Model Card Information

This GGUF repo contains quantized versions of Granite-4.2-3B, including evaluation results and comprehensive serving instructions. The quantization was performed using llama.cpp's quantization pipeline, preserving the model's reasoning capabilities while reducing size for deployment.

For the full source model documentation, including detailed training methodology, evaluation benchmarks, and advanced usage, refer to the source repo: ibm-granite/granite-4.2-3b.

---

Note: These files are optimized for llama.cpp and are not compatible with vLLM, SGLang, or the Transformers library in their current format. For those frameworks, use the source model ibm-granite/granite-4.2-3b.