FitMyLLM
AMD/ GCN 1.0

AMD Radeon HD 8970M

4 GB
VRAM (Budget)
154 GB/s
Memory Bandwidth
64 / 290
Models fit at Q4 (22%)
~20 tok/s
7B Q4 speed (Good)
Model coverage at Q422% of all models

Specifications

VRAM
4 GB
Bandwidth
154 GB/s
FP16 Compute
2.3 TFLOPS
TDP
100W
Memory
GDDR5
Architecture
GCN 1.0
Stream Processors
1,280
Compute Units
20
PCIe
Gen 3 x16
63
Fast models (>30 tok/s)
Real-time chat speed
64
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
1104
tok/s
9% VRAM
S
SmolLM2 135M0.135B
smollm2K ctxchat
1014
tok/s
9% VRAM
S
Nomic Embed Text v1.50.137B
embedding8K ctxembedding
999
tok/s
7% VRAM
S
mxbai-embed-large-v10.335B
embedding1K ctxembedding
409
tok/s
11% VRAM
S
Snowflake Arctic Embed L0.335B
embedding1K ctxembedding
409
tok/s
11% VRAM
S
GPT-2 Medium 345M0.345B
gpt21K ctxchat
397
tok/s
12% VRAM
S
SmolLM2 360M0.36B
smollm8K ctxchat
380
tok/s
12% VRAM
S
Qwen 1.5 0.5B0.5B
qwen32K ctxchat
274
tok/s
19% VRAM
S
Qwen 2.5 0.5B0.5B
qwen32K ctxchat
274
tok/s
19% VRAM
S
BGE-M30.568B
embedding8K ctxembedding
241
tok/s
15% VRAM
S
Qwen3 0.6B0.6B
qwen32K ctxchatreasoning
228
tok/s
16% VRAM
S
GPT-2 Large 774M0.774B
gpt21K ctxchat
177
tok/s
18% VRAM
S
Qwen3-0.6B0.8B
qwen40K ctxchat
171
tok/s
23% VRAM
S
Qwen3.5-0.8B0.9B
qwen256K ctxchat
152
tok/s
25% VRAM
S
Falcon3-1B1B
falcon32K ctxchat
137
tok/s
25% VRAM
S
InternLM2 1B1B
internlm32K ctxchat
137
tok/s
25% VRAM
S
TinyLlama-1.1B1.1B
llama2K ctxchat
124
tok/s
28% VRAM
S
TinyLlama 1.1B1.1B
llama2K ctxchat
124
tok/s
28% VRAM
S
Llama-3.2-1B1.2B
llama4K ctxchat
114
tok/s
29% VRAM
S
LFM2.5-1.2B-Thinking1.2B
lfm122K ctxchatreasoningtool_use
114
tok/s
18% VRAM
S
DeepSeek Coder 1.3B1.3B
deepseek16K ctxcoding
105
tok/s
30% VRAM
S
OPT 1.3B1.3B
opt2K ctxchat
105
tok/s
29% VRAM
S
Phi-1 1.3B1.3B
phi2K ctxcoding
105
tok/s
30% VRAM
S
Phi-1.5 1.3B1.3B
phi2K ctxchatcoding
105
tok/s
30% VRAM
S
GPT-2 XL 1.5B1.5B
gpt21K ctxchat
91
tok/s
32% VRAM
S
Qwen2.5-Coder-1.5B1.5B
qwen32K ctxcodingchat
91
tok/s
33% VRAM
S
Qwen2 Math 1.5B1.5B
qwen4K ctxreasoning
91
tok/s
33% VRAM
S
Qwen 2.5 1.5B1.5B
qwen32K ctxchatcoding
91
tok/s
33% VRAM
S
Yi Coder 1.5B1.5B
yi125K ctxcoding
91
tok/s
33% VRAM
S
stablelm-2-1_6b1.6B
stablelm4K ctxchat
86
tok/s
35% VRAM
S
Qwen3 1.7B1.7B
qwen32K ctxchatreasoning
81
tok/s
35% VRAM
S
SmolLM2 1.7B1.71B
smollm8K ctxchat
80
tok/s
35% VRAM
S
Qwen 1.5 1.8B1.8B
qwen32K ctxchat
76
tok/s
38% VRAM
S
Moondream2 1.9B1.9B
other2K ctxvisionchat
72
tok/s
34% VRAM
S
Gemma 1 2B2B
gemma8K ctxchat
68
tok/s
40% VRAM
S
Granite 3.0 2B2B
granite128K ctxchatcoding
68
tok/s
39% VRAM
S
Granite 3.1 2B2B
granite128K ctxchatcoding
68
tok/s
39% VRAM
S
Qwen2-VL 2B2.21B
qwen32K ctxchatvision
62
tok/s
42% VRAM
S
Qwen3.5-2B2.3B
qwen256K ctxchat
60
tok/s
45% VRAM
A
CodeGemma 2B2.51B
gemma8K ctxcoding
55
tok/s
46% VRAM

Get personalized recommendations

See ranked models with benchmark scores, run commands, and precise speed estimates for your Radeon HD 8970M.

Find Models for Radeon HD 8970MCompare with other GPUs