FitMyLLM
AMD· GCN 3.0

AMD Radeon Pro Duo

Running LLMs on the Radeon Pro Duo — the long read: which models fit at which quantisation, and the settings worth changing. · Or what a budget buys

VRAM
4 GB
BUDGET
BANDWIDTH
512
GB/S
MODELS Q4
97/449
22%
7B Q4 SPEED
~65
BLAZING
▸ MODEL COVERAGE @ Q422% OF ALL
▸ ESTIMATED SPEED· BY MODEL SIZE @ Q4

Average speeds at Q4 quantization. Actual performance varies by model architecture and context length.

3B
~152
TOK/S
7B
3.9GB NEEDED
14B
7.9GB NEEDED
32B
18.0GB NEEDED
70B
39.4GB NEEDED
▸ MEASURED RIG REPORTS

We rent the machine and time every model on it: decode, VRAM peak, concurrency, watts and cost per million tokens. Yours may already be one of them — and two are free to read in full.

SEE THE REPORTS
▸ SPECIFICATIONS
VRAM
4 GB
BANDWIDTH
512 GB/s
FP16 COMPUTE
8.2 TFLOPS
TDP
350W
MEMORY
HBM
ARCHITECTURE
GCN 3.0
STREAM PROCESSORS
4,096
COMPUTE UNITS
64
PCIE
Gen 3 x16
97
FAST MODELS · >30 TOK/S
Real-time chat speed
97
USABLE · >10 TOK/S
Comfortable for all tasks
97
TOTAL COMPATIBLE
Fit in VRAM at Q4
▸ DON’T WANT TO BUY?

Test Radeon Pro Duo (or anything bigger) without committing. Pay by the second, cancel anytime.

Spin up in ~60s. Pay by the second. Cancel anytime.

Some links are affiliate links — we may earn a small commission at no extra cost to you. This helps keep FitMyLLM free and independent.

▸ COMPATIBLE MODELS· 97
S
Falcon-H1R Tiny 90M0.09B
FALCON·256K CTX· CHAT· REASONING
5057
TOK/S · 14% VRAM
S
SmolLM2 135M0.135B
SMOLLM·2K CTX· CHAT
3371
TOK/S · 14% VRAM
S
nomic-embed-text-v1.5 100M0.14B
EMBEDDING·8K CTX· CHAT
3251
TOK/S · 14% VRAM
S
GPT-2 124M0.14B
GPT2·1K CTX· CHAT
3251
TOK/S · 14% VRAM
S
SmolVLM 256M0.256B
SMOLLM·8K CTX· CHAT· VISION
1778
TOK/S · 16% VRAM
S
Gemma 3 270M0.27B
GEMMA·32K CTX· CHAT
1686
TOK/S · 16% VRAM
S
Snowflake Arctic Embed M v2.00.305B
EMBEDDING·8K CTX· EMBEDDING· MULTILINGUAL
1492
TOK/S · 17% VRAM
S
bge-large-en-v1.5 335M0.335B
EMBEDDING·1K CTX· CHAT
1359
TOK/S · 17% VRAM
S
mxbai-embed-large-v10.335B
EMBEDDING·1K CTX· EMBEDDING
1359
TOK/S · 17% VRAM
S
Snowflake Arctic Embed L0.335B
EMBEDDING·1K CTX· EMBEDDING
1359
TOK/S · 17% VRAM
S
LFM2 350M0.35B
LFM·125K CTX· CHAT· TOOL_USE
1300
TOK/S · 18% VRAM
S
SmolLM2 360M0.36B
SMOLLM·8K CTX· CHAT
1264
TOK/S · 18% VRAM
S
GPT-2 Medium 345M0.38B
GPT2·1K CTX· CHAT
1198
TOK/S · 18% VRAM
S
Qwen 2.5 0.5B0.5B
QWEN·32K CTX· CHAT
910
TOK/S · 20% VRAM
S
SmolVLM 500M0.5B
SMOLLM·8K CTX· CHAT· VISION
910
TOK/S · 20% VRAM
S
Falcon-H1 0.5B0.52B
FALCON·128K CTX· CHAT
875
TOK/S · 20% VRAM
S
BGE-M30.568B
EMBEDDING·8K CTX· EMBEDDING
801
TOK/S · 21% VRAM
S
Snowflake Arctic Embed L v2.00.568B
EMBEDDING·8K CTX· EMBEDDING· MULTILINGUAL
801
TOK/S · 21% VRAM
S
Qwen3-Embedding 0.6B0.6B
EMBEDDING·32K CTX· EMBEDDING· MULTILINGUAL
759
TOK/S · 21% VRAM
S
Falcon-H1R Tiny 0.6B0.6B
FALCON·32K CTX· CHAT· REASONING
759
TOK/S · 21% VRAM
S
Falcon Perception 0.6B0.6B
FALCON·4K CTX· VISION
759
TOK/S · 21% VRAM
S
Qwen 1.5 0.5B0.62B
QWEN·32K CTX· CHAT
734
TOK/S · 22% VRAM
S
LFM2 700M0.74B
LFM·125K CTX· CHAT· TOOL_USE
615
TOK/S · 24% VRAM
S
Qwen3 0.6B0.75B
QWEN·32K CTX· CHAT· REASONING
607
TOK/S · 24% VRAM
S
GPT-2 Large 774M0.81B
GPT2·1K CTX· CHAT
562
TOK/S · 25% VRAM
S
Qwen 3.5 0.8B0.87B
QWEN·256K CTX· CHAT· CODING· MULTILINGUAL
523
TOK/S · 26% VRAM
S
InternVL3 1B0.94B
OTHER·32K CTX· CHAT· VISION
484
TOK/S · 27% VRAM
S
MiniCPM5 1B1.08B
MINICPM·128K CTX· CHAT· REASONING· MULTILINGUAL
421
TOK/S · 29% VRAM
S
TinyLlama 1.1B1.1B
LLAMA·2K CTX· CHAT
414
TOK/S · 29% VRAM
S
LFM2.5-1.2B-Thinking1.2B
LFM·122K CTX· CHAT· REASONING· TOOL_USE
379
TOK/S · 31% VRAM
S
Llama-3.2-1B1.2B
LLAMA·4K CTX· CHAT
379
TOK/S · 31% VRAM
S
LFM2 1.2B1.2B
LFM·125K CTX· CHAT· TOOL_USE· MULTILINGUAL
379
TOK/S · 31% VRAM
S
Zamba2 1.2B1.2B
OTHER·4K CTX· CHAT
379
TOK/S · 31% VRAM
S
EXAONE-4.0-1.2B1.3B
EXAONE·64K CTX· CHAT
350
TOK/S · 32% VRAM
S
OPT 1.3B1.3B
OPT·2K CTX· CHAT
350
TOK/S · 32% VRAM
S
MiniCPM-V 4.61.3B
OTHER·256K CTX· CHAT· VISION
350
TOK/S · 32% VRAM
S
DeepSeek Coder 1.3B1.35B
DEEPSEEK·16K CTX· CODING
337
TOK/S · 33% VRAM
S
Phi-1 1.3B1.42B
PHI·2K CTX· CODING
321
TOK/S · 34% VRAM
S
Phi-1.5 1.3B1.42B
PHI·2K CTX· CHAT· CODING
321
TOK/S · 34% VRAM
S
Qwen2.5-Coder-1.5B1.5B
QWEN·32K CTX· CHAT· TOOL_USE· CODING
303
TOK/S · 35% VRAM
▸ NEXT STEP

Get personalized recommendations.

See ranked models with benchmark scores, run commands, and precise speed estimates for your Radeon Pro Duo.

WHAT THIS CARD IS WORTH

Radeon Pro Duo holds 97 of the models in our catalogue and is, in practice, a Q4_K_M card — the largest it takes is InternLM2 5B at Q4_K_M.

TOKENS/SEC PER $100
8B at Q4_K_M, so cards compare like for like
VRAM PER $100
what memory costs on this card
THE BIGGEST IT TAKES
InternLM2 5B4.5B · Q4_K_M3.6 GB83 tok/sEST
Ministral 3 3B Reasoning4.25B · Q4_K_M3.3 GB87 tok/sEST
Qwen3 4B4B · Q4_K_M3.3 GB91 tok/sEST
Qwen3-4B Instruct 25074B · Q4_K_M3.3 GB91 tok/sEST
Qwen3-Embedding 4B4B · Q4_K_M3.3 GB91 tok/sEST
Nemotron 3 Nano 4B3.97B · Q4_K_M3.4 GB91 tok/sEST
Ministral 3 3B3.85B · Q5_K_M3.5 GB83 tok/sEST
phi-3-mini-4k 3.8B3.8B · Q4_K_M3.4 GB94 tok/sEST
AGAINST CARDS OF SIMILAR MEMORY
RTX 3070 Ti8 GB$49912.6 tok/s per $100
RTX 3060 Ti GDDR6X8 GB$39915.8 tok/s per $100
RTX 3070 Ti 8 GB GA1028 GB$59910.5 tok/s per $100
Arc A7508 GB$19910.1 tok/s per $100
Radeon Pro Duo4 GB

Neighbours in memory rather than in price: memory decides whether a card can do the job at all, so two cards of the same size at different prices is the comparison you are making. Ordered by memory, then bandwidth — not by the value column, which is worked out from bandwidth and price alone and therefore rewards a cheap card whatever its software stack does to that bandwidth in practice. Read it as one input, not as a ranking.

▸ DEVICE UNDER TEST

AMD Radeon Pro Duo — 4 GB VRAM.

RADEON PRO DUO SPEC
BRAND
AMD
VRAM
4 GB HBM
BANDWIDTH
512 GB/s
FP16 COMPUTE
8.2 TFLOPS
FP32 COMPUTE
8.2 TFLOPS
STREAM PROCESSORS
4,096
TDP
350 W
ARCHITECTURE
GCN 3.0
▸ AI CAPABILITY
97/ 449 models @ Q4

With 4 GB VRAM and 512 GB/s bandwidth, this GPU handles models up to 4.25B parameters.

Speed ≈ bandwidth / model_size × efficiency. A 7B model at Q4 runs at ~59 tok/s.

§ 01TOP MODELS FOR RADEON PRO DUO
97 FIT · SHOWING 20
MODELSIZEVRAM Q4TOK/SAVG
Ministral 3 3B Reasoning4.25B3.1 GB107
Qwen 1.5 4B4B2.9 GB11412.6
Qwen3 4B4B2.9 GB11440.7
Qwen3-4B Instruct 25074B2.9 GB11437.2
Qwen3-Embedding 4B4B2.9 GB114
Nemotron 3 Nano 4B3.97B2.9 GB11532.0
Ministral 3 3B3.85B2.8 GB11821.4
Phi-3.5 Mini 3.8B3.82B2.8 GB11946.6
phi-3-mini-4k 3.8B3.8B2.8 GB12030.5
Phi-4-mini 3.8B3.8B2.8 GB12049.0
Cogito 3B3.61B2.7 GB12622.1
Falcon3-3B3.23B2.5 GB14125.7
granite-4.0-h-micro 3.2B3.2B2.4 GB14218.4
Llama-3.2-3B3.2B2.4 GB14217.9
Falcon-H1 3B3.15B2.4 GB14449.5
Qwen 2.5 3B3.1B2.4 GB14737.2
SmolLM3-3B3.1B2.4 GB14730.5
Ministral 3B3B2.3 GB15229.6
StarCoder2 3B3B2.3 GB1529.5
Granite 4.1 3B3B2.3 GB15216.6