FitMyLLM
NVIDIA/ Blackwell 2.0

NVIDIA GeForce RTX 5090 D

32 GB
VRAM (High-end)
1790 GB/s
Memory Bandwidth
236 / 290
Models fit at Q4 (81%)
~205 tok/s
7B Q4 speed (Blazing)
Model coverage at Q481% of all models

Specifications

VRAM
32 GB
Bandwidth
1790 GB/s
FP16 Compute
104.8 TFLOPS
TDP
575W
Memory
GDDR7
Architecture
Blackwell 2.0
CUDA Cores
21,760
Tensor Cores
680
PCIe
Gen 5 x16
MSRP
$1,999
236
Fast models (>30 tok/s)
Real-time chat speed
236
Usable models (>10 tok/s)
Comfortable for all tasks
236
Total compatible
Fit in VRAM at Q4

Where to Buy

NVIDIA StoreAmazonNeweggBest Buy

MSRP: $1,999

Compatible Models(236)

S
GPT-2 124M0.124B
gpt21K ctxchat
11548
tok/s
1% VRAM
S
SmolLM2 135M0.135B
smollm2K ctxchat
10607
tok/s
1% VRAM
S
Nomic Embed Text v1.50.137B
embedding8K ctxembedding
10453
tok/s
1% VRAM
S
mxbai-embed-large-v10.335B
embedding1K ctxembedding
4275
tok/s
1% VRAM
S
Snowflake Arctic Embed L0.335B
embedding1K ctxembedding
4275
tok/s
1% VRAM
S
GPT-2 Medium 345M0.345B
gpt21K ctxchat
4151
tok/s
2% VRAM
S
SmolLM2 360M0.36B
smollm8K ctxchat
3978
tok/s
2% VRAM
S
Qwen 1.5 0.5B0.5B
qwen32K ctxchat
2864
tok/s
2% VRAM
S
Qwen 2.5 0.5B0.5B
qwen32K ctxchat
2864
tok/s
2% VRAM
S
BGE-M30.568B
embedding8K ctxembedding
2521
tok/s
2% VRAM
S
Qwen3 0.6B0.6B
qwen32K ctxchatreasoning
2387
tok/s
2% VRAM
S
GPT-2 Large 774M0.774B
gpt21K ctxchat
1850
tok/s
2% VRAM
S
Qwen3-0.6B0.8B
qwen40K ctxchat
1790
tok/s
3% VRAM
S
Qwen3.5-0.8B0.9B
qwen256K ctxchat
1591
tok/s
3% VRAM
S
Falcon3-1B1B
falcon32K ctxchat
1432
tok/s
3% VRAM
S
InternLM2 1B1B
internlm32K ctxchat
1432
tok/s
3% VRAM
S
TinyLlama-1.1B1.1B
llama2K ctxchat
1302
tok/s
3% VRAM
S
TinyLlama 1.1B1.1B
llama2K ctxchat
1302
tok/s
3% VRAM
S
Llama-3.2-1B1.2B
llama4K ctxchat
1193
tok/s
4% VRAM
S
LFM2.5-1.2B-Thinking1.2B
lfm122K ctxchatreasoningtool_use
1193
tok/s
2% VRAM
S
DeepSeek Coder 1.3B1.3B
deepseek16K ctxcoding
1102
tok/s
4% VRAM
S
OPT 1.3B1.3B
opt2K ctxchat
1102
tok/s
4% VRAM
S
Phi-1 1.3B1.3B
phi2K ctxcoding
1102
tok/s
4% VRAM
S
Phi-1.5 1.3B1.3B
phi2K ctxchatcoding
1102
tok/s
4% VRAM
S
GPT-2 XL 1.5B1.5B
gpt21K ctxchat
955
tok/s
4% VRAM
S
Qwen2.5-Coder-1.5B1.5B
qwen32K ctxcodingchat
955
tok/s
4% VRAM
S
Qwen2 Math 1.5B1.5B
qwen4K ctxreasoning
955
tok/s
4% VRAM
S
Qwen 2.5 1.5B1.5B
qwen32K ctxchatcoding
955
tok/s
4% VRAM
S
Yi Coder 1.5B1.5B
yi125K ctxcoding
955
tok/s
4% VRAM
S
stablelm-2-1_6b1.6B
stablelm4K ctxchat
895
tok/s
4% VRAM
S
Qwen3 1.7B1.7B
qwen32K ctxchatreasoning
842
tok/s
4% VRAM
S
SmolLM2 1.7B1.71B
smollm8K ctxchat
837
tok/s
4% VRAM
S
Qwen 1.5 1.8B1.8B
qwen32K ctxchat
796
tok/s
5% VRAM
S
Moondream2 1.9B1.9B
other2K ctxvisionchat
754
tok/s
4% VRAM
S
Granite 4.0 Tiny7BMoE
granite125K ctxchatcodingmultilingual
716
tok/s
14% VRAM
S
Gemma 1 2B2B
gemma8K ctxchat
716
tok/s
5% VRAM
S
Granite 3.0 2B2B
granite128K ctxchatcoding
716
tok/s
5% VRAM
S
Granite 3.1 2B2B
granite128K ctxchatcoding
716
tok/s
5% VRAM
S
Qwen2-VL 2B2.21B
qwen32K ctxchatvision
648
tok/s
5% VRAM
S
Qwen3.5-2B2.3B
qwen256K ctxchat
623
tok/s
6% VRAM

Get personalized recommendations

See ranked models with benchmark scores, run commands, and precise speed estimates for your GeForce RTX 5090 D.

Find Models for GeForce RTX 5090 DCompare with other GPUs