FitMyLLM
INTEL/ Generation 12.1

Intel Xe DG1

4 GB
VRAM (Budget)
68 GB/s
Memory Bandwidth
64 / 290
Models fit at Q4 (22%)
~3 tok/s
7B Q4 speed (Slow)
Model coverage at Q422% of all models

Specifications

VRAM
4 GB
Bandwidth
68 GB/s
FP16 Compute
4 TFLOPS
TDP
30W
Memory
LPDDR4X
Architecture
Generation 12.1
Compute Units
40
PCIe
Gen 4 x8
11
Fast models (>30 tok/s)
Real-time chat speed
38
Usable models (>10 tok/s)
Comfortable for all tasks
64
Total compatible
Fit in VRAM at Q4

Compatible Models(64)

S
GPT-2 124M0.124B
gpt21K ctxchat
175
tok/s
9% VRAM
S
SmolLM2 135M0.135B
smollm2K ctxchat
161
tok/s
9% VRAM
S
Nomic Embed Text v1.50.137B
embedding8K ctxembedding
159
tok/s
7% VRAM
S
mxbai-embed-large-v10.335B
embedding1K ctxembedding
65
tok/s
11% VRAM
S
Snowflake Arctic Embed L0.335B
embedding1K ctxembedding
65
tok/s
11% VRAM
S
GPT-2 Medium 345M0.345B
gpt21K ctxchat
63
tok/s
12% VRAM
S
SmolLM2 360M0.36B
smollm8K ctxchat
60
tok/s
12% VRAM
A
Qwen 1.5 0.5B0.5B
qwen32K ctxchat
44
tok/s
19% VRAM
A
Qwen 2.5 0.5B0.5B
qwen32K ctxchat
44
tok/s
19% VRAM
B
BGE-M30.568B
embedding8K ctxembedding
38
tok/s
15% VRAM
B
Qwen3 0.6B0.6B
qwen32K ctxchatreasoning
36
tok/s
16% VRAM
B
GPT-2 Large 774M0.774B
gpt21K ctxchat
28
tok/s
18% VRAM
B
Qwen3-0.6B0.8B
qwen40K ctxchat
27
tok/s
23% VRAM
C
Qwen3.5-0.8B0.9B
qwen256K ctxchat
24
tok/s
25% VRAM
C
Falcon3-1B1B
falcon32K ctxchat
22
tok/s
25% VRAM
C
InternLM2 1B1B
internlm32K ctxchat
22
tok/s
25% VRAM
C
TinyLlama-1.1B1.1B
llama2K ctxchat
20
tok/s
28% VRAM
C
TinyLlama 1.1B1.1B
llama2K ctxchat
20
tok/s
28% VRAM
C
Llama-3.2-1B1.2B
llama4K ctxchat
18
tok/s
29% VRAM
C
LFM2.5-1.2B-Thinking1.2B
lfm122K ctxchatreasoningtool_use
18
tok/s
18% VRAM
C
DeepSeek Coder 1.3B1.3B
deepseek16K ctxcoding
17
tok/s
30% VRAM
C
OPT 1.3B1.3B
opt2K ctxchat
17
tok/s
29% VRAM
C
Phi-1 1.3B1.3B
phi2K ctxcoding
17
tok/s
30% VRAM
C
Phi-1.5 1.3B1.3B
phi2K ctxchatcoding
17
tok/s
30% VRAM
C
GPT-2 XL 1.5B1.5B
gpt21K ctxchat
15
tok/s
32% VRAM
C
Qwen2.5-Coder-1.5B1.5B
qwen32K ctxcodingchat
15
tok/s
33% VRAM
C
Qwen2 Math 1.5B1.5B
qwen4K ctxreasoning
15
tok/s
33% VRAM
C
Qwen 2.5 1.5B1.5B
qwen32K ctxchatcoding
15
tok/s
33% VRAM
C
Yi Coder 1.5B1.5B
yi125K ctxcoding
15
tok/s
33% VRAM
D
stablelm-2-1_6b1.6B
stablelm4K ctxchat
14
tok/s
35% VRAM
D
SmolLM2 1.7B1.71B
smollm8K ctxchat
13
tok/s
35% VRAM
D
Qwen3 1.7B1.7B
qwen32K ctxchatreasoning
13
tok/s
35% VRAM
D
Qwen 1.5 1.8B1.8B
qwen32K ctxchat
12
tok/s
38% VRAM
D
Gemma 1 2B2B
gemma8K ctxchat
11
tok/s
40% VRAM
D
Granite 3.0 2B2B
granite128K ctxchatcoding
11
tok/s
39% VRAM
D
Granite 3.1 2B2B
granite128K ctxchatcoding
11
tok/s
39% VRAM
D
Moondream2 1.9B1.9B
other2K ctxvisionchat
11
tok/s
34% VRAM
D
Qwen2-VL 2B2.21B
qwen32K ctxchatvision
10
tok/s
42% VRAM
D
CodeGemma 2B2.51B
gemma8K ctxcoding
9
tok/s
46% VRAM
D
Qwen3.5-2B2.3B
qwen256K ctxchat
9
tok/s
45% VRAM

Get personalized recommendations

See ranked models with benchmark scores, run commands, and precise speed estimates for your Xe DG1.

Find Models for Xe DG1Compare with other GPUs