Skip to content

Gnani Evon V3.3 30B A3B VRAM requirements

How much VRAM Gnani Evon V3.3 30B A3B needs
4-bit weights: 17.9 GBWith 8K context: 18.9 GBWith 32K context: 19.0 GB8-bit, 8K context: 32.4 GB
Parameters
31.7B
Active per token
3B
Longest context
128K
Released
Aug 2026
License
Apache 2.0
Strongest at
Not rated

How much VRAM Gnani Evon V3.3 30B A3B needs

Memory Gnani Evon V3.3 30B A3B needs at each precision and context length
PrecisionWeightsWith 8KWith 32KWith 128K
16-bit59.1 GB60.1 GB60.2 GB60.8 GB
8-bit31.4 GB32.4 GB32.5 GB33.1 GB
6-bit24.2 GB25.2 GB25.3 GB25.9 GB
5-bit21.0 GB22.0 GB22.1 GB22.7 GB
4-bit17.9 GB18.9 GB19.0 GB19.6 GB
3-bit14.5 GB15.4 GB15.5 GB16.1 GB
2-bit11.1 GB12.0 GB12.2 GB12.7 GB
1.6-bit6.1 GB7.0 GB7.2 GB7.7 GB

One conversation, 16-bit context cache, runtime overhead included.

GPUs and Macs that run Gnani Evon V3.3 30B A3B

The best way Gnani Evon V3.3 30B A3B runs on common GPUs, Macs and AI boxes
MachineBest fitContextSpeedFirst replyReport
RTX 4060Too bigPDF: Gnani Evon V3.3 30B A3B on RTX 4060 report
RTX 5060Too bigPDF: Gnani Evon V3.3 30B A3B on RTX 5060 report
RTX 3070Too bigPDF: Gnani Evon V3.3 30B A3B on RTX 3070 report
RTX 3080 10 GBToo bigPDF: Gnani Evon V3.3 30B A3B on RTX 3080 10 GB report
RTX 3060 12 GBToo bigPDF: Gnani Evon V3.3 30B A3B on RTX 3060 12 GB report
Arc B580Too bigPDF: Gnani Evon V3.3 30B A3B on Arc B580 report
RTX 4070Too bigPDF: Gnani Evon V3.3 30B A3B on RTX 4070 report
RTX 4070 SuperToo bigPDF: Gnani Evon V3.3 30B A3B on RTX 4070 Super report
RTX 5070Too bigPDF: Gnani Evon V3.3 30B A3B on RTX 5070 report
RTX 4060 Ti 16 GBToo bigPDF: Gnani Evon V3.3 30B A3B on RTX 4060 Ti 16 GB report
RTX 5060 Ti 16 GBToo bigPDF: Gnani Evon V3.3 30B A3B on RTX 5060 Ti 16 GB report
Arc A770 16 GBToo bigPDF: Gnani Evon V3.3 30B A3B on Arc A770 16 GB report
Radeon RX 9070 XTToo bigPDF: Gnani Evon V3.3 30B A3B on Radeon RX 9070 XT report
RTX 4070 Ti SuperToo bigPDF: Gnani Evon V3.3 30B A3B on RTX 4070 Ti Super report
RTX 4080 SuperToo bigPDF: Gnani Evon V3.3 30B A3B on RTX 4080 Super report
RTX 5070 TiToo bigPDF: Gnani Evon V3.3 30B A3B on RTX 5070 Ti report
RTX 5080Too bigPDF: Gnani Evon V3.3 30B A3B on RTX 5080 report
Radeon RX 7900 XTX5-bit128K118 tok/s0.8 sPDF: Gnani Evon V3.3 30B A3B on Radeon RX 7900 XTX report
RTX 30905-bit128K155 tok/s0.5 sPDF: Gnani Evon V3.3 30B A3B on RTX 3090 report
RTX 40905-bit128K161 tok/s0.2 sPDF: Gnani Evon V3.3 30B A3B on RTX 4090 report
Radeon AI PRO R97006-bit128K89 tok/s0.5 sPDF: Gnani Evon V3.3 30B A3B on Radeon AI PRO R9700 report
RTX 50906-bit128K195 tok/s0.2 sPDF: Gnani Evon V3.3 30B A3B on RTX 5090 report
2× RTX 30908-bit128K124 tok/s0.3 sPDF: Gnani Evon V3.3 30B A3B on 2× RTX 3090 report
2× RTX 40908-bit128K130 tok/s0.1 sPDF: Gnani Evon V3.3 30B A3B on 2× RTX 4090 report
RTX PRO 5000 Blackwell 48 GB8-bit128K152 tok/s0.3 sPDF: Gnani Evon V3.3 30B A3B on RTX PRO 5000 Blackwell 48 GB report
RTX 6000 Ada8-bit128K126 tok/s0.2 sPDF: Gnani Evon V3.3 30B A3B on RTX 6000 Ada report
L40S8-bit128K118 tok/s0.3 sPDF: Gnani Evon V3.3 30B A3B on L40S report
2× RTX 50908-bit128K175 tok/s0.1 sPDF: Gnani Evon V3.3 30B A3B on 2× RTX 5090 report
4× RTX 30908-bit128K124 tok/s0.2 sPDF: Gnani Evon V3.3 30B A3B on 4× RTX 3090 report
RTX PRO 6000 Blackwell8-bit128K175 tok/s0.1 sPDF: Gnani Evon V3.3 30B A3B on RTX PRO 6000 Blackwell report
Mac mini M4 16 GBToo bigPDF: Gnani Evon V3.3 30B A3B on Mac mini M4 16 GB report
Mac mini M4 24 GBToo bigPDF: Gnani Evon V3.3 30B A3B on Mac mini M4 24 GB report
Mac mini M4 32 GB4-bit128K38 tok/s11 sPDF: Gnani Evon V3.3 30B A3B on Mac mini M4 32 GB report
Mac mini M4 Pro 24 GBToo bigPDF: Gnani Evon V3.3 30B A3B on Mac mini M4 Pro 24 GB report
Mac mini M4 Pro 48 GB8-bit128K46 tok/s5.5 sPDF: Gnani Evon V3.3 30B A3B on Mac mini M4 Pro 48 GB report
Mac mini M4 Pro 64 GB8-bit128K46 tok/s5.5 sPDF: Gnani Evon V3.3 30B A3B on Mac mini M4 Pro 64 GB report
M4 Max 128 GB8-bit128K66 tok/s2.8 sPDF: Gnani Evon V3.3 30B A3B on M4 Max 128 GB report
M5 Max 128 GB8-bit128K70 tok/s0.8 sPDF: Gnani Evon V3.3 30B A3B on M5 Max 128 GB report
M3 Ultra 96 GB8-bit128K75 tok/s1.5 sPDF: Gnani Evon V3.3 30B A3B on M3 Ultra 96 GB report
M3 Ultra 256 GB8-bit128K75 tok/s1.5 sPDF: Gnani Evon V3.3 30B A3B on M3 Ultra 256 GB report
M3 Ultra 512 GB8-bit128K75 tok/s1.5 sPDF: Gnani Evon V3.3 30B A3B on M3 Ultra 512 GB report
M5 Ultra 512 GB8-bit128K87 tok/s0.5 sPDF: Gnani Evon V3.3 30B A3B on M5 Ultra 512 GB report
Ryzen AI Max+ 395 64 GB8-bit128K39 tok/s1.6 sPDF: Gnani Evon V3.3 30B A3B on Ryzen AI Max+ 395 64 GB report
Ryzen AI Max+ 395 128 GB8-bit128K39 tok/s1.6 sPDF: Gnani Evon V3.3 30B A3B on Ryzen AI Max+ 395 128 GB report
RTX Spark8-bit128K54 tok/s0.6 sPDF: Gnani Evon V3.3 30B A3B on RTX Spark report
DGX Spark8-bit128K50 tok/s0.7 sPDF: Gnani Evon V3.3 30B A3B on DGX Spark report
Dual DGX Spark8-bit128K50 tok/s0.4 sPDF: Gnani Evon V3.3 30B A3B on Dual DGX Spark report
DGX Station GB3008-bit128K265 tok/s0.1 sPDF: Gnani Evon V3.3 30B A3B on DGX Station GB300 report
2× RTX PRO 6000 Blackwell8-bit128K175 tok/s0.1 sPDF: Gnani Evon V3.3 30B A3B on 2× RTX PRO 6000 Blackwell report
4× RTX PRO 6000 Blackwell8-bit128K175 tok/s0.1 sPDF: Gnani Evon V3.3 30B A3B on 4× RTX PRO 6000 Blackwell report
8× RTX PRO 6000 Blackwell8-bit128K175 tok/s0.1 sPDF: Gnani Evon V3.3 30B A3B on 8× RTX PRO 6000 Blackwell report
A100 80 GB SXM8-bit128K186 tok/s0.4 sPDF: Gnani Evon V3.3 30B A3B on A100 80 GB SXM report
H100 SXM8-bit128K222 tok/s0.1 sPDF: Gnani Evon V3.3 30B A3B on H100 SXM report
4× H200 NVL8-bit128K245 tok/s0.1 sPDF: Gnani Evon V3.3 30B A3B on 4× H200 NVL report
8× H200 NVL8-bit128K245 tok/s0.1 sPDF: Gnani Evon V3.3 30B A3B on 8× H200 NVL report

Best fit is the highest quality, up to 8-bit, that leaves 16K of context. First reply is the wait for a 2K prompt. Speeds are rough, plus or minus 25%.

How to run it

llama.cpp
llama-server -hf luminenio/gnani-evon-v3.3-30B-A3B-GGUF
vLLM
vllm serve gnani/gnani-evon-v3.3-30B-A3B

Questions

How much VRAM does Gnani Evon V3.3 30B A3B need?

Gnani Evon V3.3 30B A3B needs 17.9 GB for its weights at 4-bit, 18.9 GB with an 8K context and 19.0 GB with 32K. At 8-bit, 32.4 GB.

What is the minimum VRAM for Gnani Evon V3.3 30B A3B?

About 15.4 GB at 3-bit with an 8K context, the lowest precision we recommend for a model this size. The smallest machine here that runs it without spilling is the Radeon RX 7900 XTX.

How big is Gnani Evon V3.3 30B A3B?

31.7B parameters (3B active per token): 59.1 GB of weights at 16-bit, 17.9 GB at 4-bit.

Can Gnani Evon V3.3 30B A3B run on an RTX 4090?

Yes. At 5-bit it holds 128K of context and writes about 161 tokens a second.

More models