Skip to content

Granite 3.3 8B

by IBM · granite family

8B

parameters

text-generation code-generation reasoning math tool-use summarization

Granite 3.3 8B is IBM's enterprise-focused language model built on a dense transformer decoder architecture with GQA, RoPE, and SwiGLU. It supports a 128K context window and features structured reasoning through dedicated thinking and response tags. The model delivers strong results on AlpacaEval-2.0 and Arena-Hard benchmarks, with particular strengths in mathematics, coding, and instruction adherence. Released under the Apache 2.0 license, it is well-suited for enterprise RAG applications and business use cases.

Quick Start with Ollama

ollama run 8b-instruct-q8_0
Resources Ollama Hugging Face Official Page
Creator IBM
Parameters 8B
Architecture transformer-decoder
Context 125K tokens
Released Apr 16, 2025
License Apache 2.0
Ollama granite3.3

Quantization Options

Format File Size VRAM Required Quality Ollama Tag
Q4_K_M 4.1 GB 6 GB 8b-instruct-q4_K_M
Q8_0 rec 7.2 GB 10 GB 8b-instruct-q8_0
F16 15.2 GB 18 GB 8b-instruct-fp16

Compatible Hardware

Q8_0 requires 10 GB VRAM

Compatible Hardware

HardwareVRAMTypeFitEst. Speed
Mac Studio M4 Ultra 512GB512 GBmacRuns~82 tok/s
Mac Pro M2 Ultra 192GB192 GBmacRuns~80 tok/s
Mac Studio M4 Ultra 192GB192 GBmacRuns~82 tok/s
Mac Studio M4 Max 128GB128 GBmacRuns~55 tok/s
MacBook Pro M4 Max 128GB128 GBmacRuns~55 tok/s
MacBook Pro M5 Max 128GB128 GBmacRuns~55 tok/s
NVIDIA RTX PRO 6000 Blackwell96 GBgpuRuns~192 tok/s
MacBook Pro M3 Max 96GB96 GBmacRuns~40 tok/s
Mac mini M4 Pro 64GB64 GBmacRuns~27 tok/s
Mac Studio M4 Max 64GB64 GBmacRuns~55 tok/s
MacBook Pro M4 Max 64GB64 GBmacRuns~55 tok/s
MacBook Pro M5 Max 64GB64 GBmacRuns~55 tok/s
NVIDIA RTX 6000 Ada Generation48 GBgpuRuns~96 tok/s
NVIDIA RTX A600048 GBgpuRuns~77 tok/s
NVIDIA RTX PRO 5000 Blackwell48 GBgpuRuns~96 tok/s
Mac mini M4 Pro 48GB48 GBmacRuns~27 tok/s
MacBook Pro M3 Max 48GB48 GBmacRuns~40 tok/s
MacBook Pro M4 Max 48GB48 GBmacRuns~55 tok/s
MacBook Pro M4 Pro 48GB48 GBmacRuns~27 tok/s
MacBook Pro M5 Max 48GB48 GBmacRuns~41 tok/s
MacBook Pro M5 Pro 48GB48 GBmacRuns~27 tok/s
Mac Studio M4 Max 36GB36 GBmacRuns~55 tok/s
MacBook Pro M3 Pro 36GB36 GBmacRuns~15 tok/s
MacBook Pro M5 Max 36GB36 GBmacRuns~41 tok/s
NVIDIA RTX 5000 Ada Generation32 GBgpuRuns~72 tok/s
NVIDIA GeForce RTX 509032 GBgpuRuns~179 tok/s
iMac M4 32GB32 GBmacRuns~12 tok/s
Mac mini M4 32GB32 GBmacRuns~12 tok/s
MacBook Air M5 32GB32 GBmacRuns~12 tok/s
MacBook Air M4 32GB32 GBmacRuns~12 tok/s
MacBook Pro M5 32GB32 GBmacRuns~12 tok/s
AMD Radeon RX 7900 XTX24 GBgpuRuns~96 tok/s
NVIDIA GeForce RTX 3090 Ti24 GBgpuRuns~101 tok/s
NVIDIA GeForce RTX 309024 GBgpuRuns~94 tok/s
NVIDIA GeForce RTX 409024 GBgpuRuns~101 tok/s
NVIDIA RTX A500024 GBgpuRuns~77 tok/s
iMac M3 24GB24 GBmacRuns~10 tok/s
Mac mini M2 24GB24 GBmacRuns~10 tok/s
Mac mini M4 Pro 24GB24 GBmacRuns~27 tok/s
MacBook Air M2 24GB24 GBmacRuns~10 tok/s
MacBook Air M4 24GB24 GBmacRuns~12 tok/s
MacBook Air M5 24GB24 GBmacRuns~12 tok/s
MacBook Pro M4 Pro 24GB24 GBmacRuns~27 tok/s
MacBook Pro M5 24GB24 GBmacRuns~12 tok/s
MacBook Pro M5 Pro 24GB24 GBmacRuns~27 tok/s
AMD Radeon RX 7900 XT20 GBgpuRuns~80 tok/s
NVIDIA RTX 4000 Ada Generation20 GBgpuRuns~36 tok/s
MacBook Pro M3 Pro 18GB18 GBmacRuns~15 tok/s
AMD Radeon RX 6800 XT16 GBgpuRuns~51 tok/s
AMD Radeon RX 6900 XT16 GBgpuRuns~51 tok/s
AMD Radeon RX 7800 XT16 GBgpuRuns~62 tok/s
AMD Radeon RX 9060 XT 16GB16 GBgpuRuns~54 tok/s
AMD Radeon RX 9070 XT16 GBgpuRuns~65 tok/s
AMD Radeon RX 907016 GBgpuRuns~54 tok/s
Intel Arc A77016 GBgpuRuns~56 tok/s
NVIDIA GeForce RTX 4060 Ti 16GB16 GBgpuRuns~29 tok/s
NVIDIA GeForce RTX 4070 Ti Super16 GBgpuRuns~67 tok/s
NVIDIA GeForce RTX 4080 Super16 GBgpuRuns~74 tok/s
NVIDIA GeForce RTX 408016 GBgpuRuns~72 tok/s
NVIDIA GeForce RTX 5060 Ti 16GB16 GBgpuRuns~45 tok/s
NVIDIA GeForce RTX 5070 Ti16 GBgpuRuns~90 tok/s
NVIDIA GeForce RTX 508016 GBgpuRuns~96 tok/s
NVIDIA RTX A400016 GBgpuRuns~45 tok/s
iMac M1 16GB16 GBmacRuns~7 tok/s
iMac M4 16GB16 GBmacRuns~12 tok/s
Mac mini M1 16GB16 GBmacRuns~7 tok/s
Mac mini M4 16GB16 GBmacRuns~12 tok/s
MacBook Air M2 16GB16 GBmacRuns~10 tok/s
MacBook Air M4 16GB16 GBmacRuns~12 tok/s
MacBook Air M3 16GB16 GBmacRuns~10 tok/s
MacBook Air M5 16GB16 GBmacRuns~12 tok/s
MacBook Pro M2 Pro 16GB16 GBmacRuns~20 tok/s
MacBook Pro M1 16GB16 GBmacRuns~7 tok/s
MacBook Pro M5 16GB16 GBmacRuns~12 tok/s
AMD Radeon RX 6700 XT12 GBgpuRuns~38 tok/s
AMD Radeon RX 7700 XT12 GBgpuRuns~43 tok/s
Intel Arc B58012 GBgpuRuns~46 tok/s
NVIDIA GeForce RTX 3060 12GB12 GBgpuRuns~36 tok/s
NVIDIA GeForce RTX 3080 12GB12 GBgpuRuns~91 tok/s
NVIDIA GeForce RTX 4070 Super12 GBgpuRuns~50 tok/s
NVIDIA GeForce RTX 4070 Ti12 GBgpuRuns~50 tok/s
NVIDIA GeForce RTX 407012 GBgpuRuns~50 tok/s
NVIDIA GeForce RTX 507012 GBgpuRuns~67 tok/s
NVIDIA GeForce GTX 1080 Ti11 GBgpuRuns (tight)~48 tok/s
NVIDIA GeForce RTX 2080 Ti11 GBgpuRuns (tight)~62 tok/s
Intel Arc B57010 GBgpuCPU Offload~11 tok/s
NVIDIA GeForce RTX 3080 10GB10 GBgpuCPU Offload~23 tok/s
AMD Radeon RX 6600 XT8 GBgpuCPU Offload~8 tok/s
AMD Radeon RX 76008 GBgpuCPU Offload~9 tok/s
AMD Radeon RX 9060 XT 8GB8 GBgpuCPU Offload~8 tok/s
Intel Arc A7508 GBgpuCPU Offload~15 tok/s
NVIDIA GeForce GTX 10708 GBgpuCPU Offload~8 tok/s
NVIDIA GeForce RTX 2060 Super8 GBgpuCPU Offload~14 tok/s
NVIDIA GeForce RTX 2070 Super8 GBgpuCPU Offload~14 tok/s
NVIDIA GeForce RTX 2080 Super8 GBgpuCPU Offload~15 tok/s
NVIDIA GeForce RTX 30508 GBgpuCPU Offload~7 tok/s
NVIDIA GeForce RTX 3060 Ti8 GBgpuCPU Offload~14 tok/s
NVIDIA GeForce RTX 30708 GBgpuCPU Offload~14 tok/s
NVIDIA GeForce RTX 4060 Ti 8GB8 GBgpuCPU Offload~9 tok/s
NVIDIA GeForce RTX 40608 GBgpuCPU Offload~8 tok/s
NVIDIA GeForce RTX 50508 GBgpuCPU Offload~7 tok/s
NVIDIA GeForce RTX 5060 Ti 8GB8 GBgpuCPU Offload~14 tok/s
NVIDIA GeForce RTX 50608 GBgpuCPU Offload~10 tok/s
MacBook Air M1 8GB8 GBmacCPU Offload~2 tok/s
MacBook Air M2 8GB8 GBmacCPU Offload~3 tok/s
2 hardware device(s) cannot run this model at Q8_0.