readygpu

granite vision 4.1 4b

4B parameters · 128k context · Apache 2.0 · released 2026-04-29

vision

In short

granite vision 4.1 4b is an open model from IBM with 4B parameters. At Q4_K_M, granite vision 4.1 4b needs about 3 GB of memory at 8k context and 3 GB at 8k, counting the weights, the KV cache and runtime buffers. The cheapest hardware in Canada that runs granite vision 4.1 4b fully in memory at 8k context is the NVIDIA GeForce RTX 3060 12GB, $497 at Amazon, at about 52 tok/s (estimated).

Updated . Speeds are estimates; how they are worked out.

What to run it on in Canada

Fully in memory at 8k context, priced at retailers today.

All hardware that runs it

Everything we price, best value first. Two cards split the model by layers.

HardwareQuantFitsDecodePricetok/s per 1,000
NVIDIA GeForce RTX 3060 12GBQ8_0fits52 tok/s$497104.6
Intel Arc B580Q8_0fits53 tok/s$56094.8
Intel Arc A770 16GBQ8_0fits65 tok/s$65599.6
NVIDIA GeForce RTX 3080 10GBQ8_0fits110 tok/s$769142.8
AMD Radeon RX 9060 XT 16GBQ8_0fits41 tok/s$80051.3
AMD Radeon RX 9070 XTQ8_0fits82 tok/s$1,05078.1
Intel Arc Pro B60Q8_0fits53 tok/s$1,06050.1
NVIDIA GeForce RTX 5060 Ti 16GBQ8_0fits65 tok/s$1,12057.8
NVIDIA GeForce RTX 5070Q8_0fits97 tok/s$1,18182.2
AMD Radeon RX 7800 XTQ8_0fits80 tok/s$1,39757.3
Mac mini M4 16GBQ8_0fits18 tok/s$1,39913.0
NVIDIA GeForce RTX 4070 SuperQ8_0fits73 tok/s$1,40052.0
AMD Radeon RX 7600 XTQ8_0fits37 tok/s$1,44025.6
NVIDIA GeForce RTX 5070 TiQ8_0fits130 tok/s$1,69076.6
NVIDIA GeForce RTX 4070 Ti SuperQ8_0fits97 tok/s$1,79054.2
NVIDIA GeForce RTX 4080 SuperQ8_0fits106 tok/s$2,00053.2
AMD Radeon RX 7900 XTQ8_0fits103 tok/s$2,08349.2
NVIDIA GeForce RTX 5080Q8_0fits139 tok/s$2,37058.5
NVIDIA GeForce RTX 3090Q8_0fits135 tok/s$2,80048.3
AMD Radeon AI PRO R9700Q8_0fits82 tok/s$2,80029.3
NVIDIA RTX PRO 6000 BlackwellQ8_0fits259 tok/s$3,13382.6
NVIDIA GeForce RTX 3090 TiQ8_0fits146 tok/s$3,99936.4
GMKtec EVO-X2 Ryzen AI Max+ 395 128GBQ8_0fits33 tok/s$5,0996.4
NVIDIA GeForce RTX 4090Q8_0fits146 tok/s$7,10720.5
ASUS Ascent GX10 128GBQ8_0fits39 tok/s$8,2994.7
NVIDIA GeForce RTX 5090Q8_0fits259 tok/s$10,30025.1
Mac Studio M4 Max 128GBQ8_0fits83 tok/s––
Mac mini M4 Pro 64GBQ8_0fits41 tok/s––
AMD Radeon RX 7900 XTXQ8_0fits123 tok/s––
Framework Desktop Ryzen AI Max+ 395 128GBQ8_0fits33 tok/s––
Mac Studio M3 Ultra 512GBQ8_0fits124 tok/s––
NVIDIA DGX Spark 128GBQ8_0fits39 tok/s––
Mac Studio M3 Ultra 256GBQ8_0fits124 tok/s––
Mac mini M4 Pro 48GBQ8_0fits41 tok/s––
NVIDIA GeForce RTX 4060 Ti 16GBQ8_0fits42 tok/s––
Mac Studio M4 Max 64GBQ8_0fits83 tok/s––
Mac Studio M3 Ultra 96GBQ8_0fits124 tok/s––
Framework Desktop Ryzen AI Max+ 395 64GBQ8_0fits33 tok/s––
Mac mini M4 32GBQ8_0fits18 tok/s––
NVIDIA Tesla P40Q8_0fits50 tok/s––

Memory needed

Weights, KV cache and runtime buffers at each context length (KV cache at fp16).

QuantFile@8k@32k@64k@128k
Q4_K_Mgguf2.0 GB3.3 GB5.3 GB8.0 GB13.4 GB
Q5_K_Mgguf2.3 GB3.6 GB5.7 GB8.3 GB13.7 GB
Q8_0gguf3.4 GB4.7 GB6.8 GB9.4 GB14.8 GB

Run it

Copy a command for the runtime you use.

  • Q4_K_M
    • Ollamaollama run hf.co/ibm-granite/granite-vision-4.1-4b-GGUF:Q4_K_M
    • LM Studiolms get ibm-granite/granite-vision-4.1-4b-GGUF@q4_k_m
    • llama.cppllama-server -hf ibm-granite/granite-vision-4.1-4b-GGUF:Q4_K_M
  • Q5_K_M
    • Ollamaollama run hf.co/ibm-granite/granite-vision-4.1-4b-GGUF:Q5_K_M
    • LM Studiolms get ibm-granite/granite-vision-4.1-4b-GGUF@q5_k_m
    • llama.cppllama-server -hf ibm-granite/granite-vision-4.1-4b-GGUF:Q5_K_M
  • Q8_0
    • Ollamaollama run hf.co/ibm-granite/granite-vision-4.1-4b-GGUF:Q8_0
    • LM Studiolms get ibm-granite/granite-vision-4.1-4b-GGUF@q8_0
    • llama.cppllama-server -hf ibm-granite/granite-vision-4.1-4b-GGUF:Q8_0

Notes

Specialised for chart, table and key-value extraction from documents; may not generalise well to open-ended vision tasks and is English-only. Requires transformers>=5.8.0 (or trust_remote_code on older versions) or vLLM>=0.21.0.

Questions

How much VRAM does granite vision 4.1 4b need?
At Q4_K_M, granite vision 4.1 4b needs about 3 GB of memory at 8k context and 3 GB at 8k, counting the weights, the KV cache and runtime buffers.
What is the best GPU for running granite vision 4.1 4b locally?
The cheapest hardware in Canada that runs granite vision 4.1 4b fully in memory at 8k context is the NVIDIA GeForce RTX 3060 12GB, $497 at Amazon, at about 52 tok/s (estimated). The best value is the NVIDIA GeForce RTX 3080 10GB (about 110 tok/s for $769). The fastest we price is the NVIDIA RTX PRO 6000 Blackwell, at about 259 tok/s.
Can I run granite vision 4.1 4b on a 16 GB GPU?
Yes: at Q8_0 it fits fully in 16 GB at 8k context. On the NVIDIA GeForce RTX 5080 it runs at about 139 tok/s (estimated).
Can I run granite vision 4.1 4b on a 24 GB GPU?
Yes: at Q8_0 it fits fully in 24 GB at 8k context. On the NVIDIA GeForce RTX 3090 Ti it runs at about 146 tok/s (estimated).
How do I run granite vision 4.1 4b locally?
With Ollama: ollama run hf.co/ibm-granite/granite-vision-4.1-4b-GGUF:Q4_K_M (the Q4_K_M build). LM Studio and llama.cpp commands are on this page.
What licence is granite vision 4.1 4b released under?
Apache 2.0, which allows commercial use.