NVIDIA A40 48GB

36B31.9 GB25 tok/s77K ctx

Hermes 4.3 36B

C52

👁 HuggingFace
SmolLM3 3B

C51

3B9.5 GB42 tok/s128K ctx

dense

👁 DeepSeek
DeepSeek R1 1.5B

C51

1.5B7.0 GB21 tok/s33K ctx

dense

👁 Alibaba
Qwen 2.5 Math 7B

C51

7B10.8 GB98 tok/s4K ctx

dense

👁 TinyLlama
TinyLlama 1.1B

C51

1.1B6.7 GB15 tok/s4K ctx

dense

EXAONE 4.0 32B

aya expanse 32b heretic MPOA i1

Baichuan M2 32B Q4 K M

baichuan inc Baichuan M2 32B

27B25.3 GB33 tok/s131K ctx

👁 LMSYS
Vicuna 7B

C50

7B17.8 GB98 tok/s4K ctx

dense

👁 Unsloth
Qwen3.5 27B

C50

27B25.3 GB33 tok/s131K ctx

dense

👁 Unsloth
gemma 3 27b it

C50

27B25.3 GB33 tok/s131K ctx

dense

gemma 3 27b it

C50

cognitivecomputations Dolphin Mistral 24B Venice Edition

👁 BigCode
StarCoder2 15B

C50

15B17.7 GB56 tok/s16K ctx

dense

👁 Unsloth
Mistral Small 3.2 24B Instruct 2506

C49

24B23.2 GB37 tok/s157K ctx

dense

👁 Alibaba
Qwen 2.5 Math 72B

C49

72B54.5 GB8 tok/s4K ctx

dense

Mistral Small 24B Instruct 2501

mistral small 3.1 24b instruct 2503 hf

cognitivecomputations Dolphin3.0 R1 Mistral 24B

Dolphin Mistral GLM 4.7 Flash 24B Venice Edition Thinking Uncensored i1

cognitivecomputations Dolphin Mistral 24B Venice Edition

👁 OpenChat
OpenChat 7B

C49

7B11.9 GB98 tok/s8K ctx

dense

👁 Cohere
Aya Expanse 8B

C49

8B12.5 GB112 tok/s8K ctx

dense

Codestral 22B v0.1

Codestral 22B v0.1

Codestral 22B v0.1

Qwen3 48B A4B Savant Commander Distill 12X Closed Open Heretic Uncensored

48B40.6 GB19 tok/s37K ctx

Codestral 22B v0.1 IMat

👁 Alibaba
Qwen 2.5 1.5B

C48

1.5B7.0 GB21 tok/s131K ctx

dense

Codestral 22B v0.1 i1

Codestral 21B Pruned i1

21B21.0 GB42 tok/s192K ctx

internlm2 5 20b chat

20B20.2 GB45 tok/s205K ctx

internlm2 limarp chat 20b

20B20.2 GB45 tok/s205K ctx

internlm2 math plus 20b i1

👁 Cognitive Computations
Dolphin 2.9 8B

C48

8B12.5 GB112 tok/s33K ctx

dense

20B20.2 GB45 tok/s205K ctx

Codestral RAG 19B Pruned i1

👁 Teknium
OpenHermes 2.5 7B

C48

7B11.9 GB98 tok/s8K ctx

dense

👁 DeepSeek
DeepSeek LLM 7B

C48

7B17.3 GB98 tok/s4K ctx

dense

👁 Unsloth
DeepSeek R1 Distill Qwen 14B

C48

14B15.9 GB64 tok/s329K ctx

dense

19B19.5 GB47 tok/s221K ctx

Phi 4 reasoning vision 15B

StarCoder2 15B

👁 Hugging Face H4
Zephyr 7B Beta

C48

7B11.9 GB98 tok/s33K ctx

dense

gemma 3 12b it

12B14.4 GB74 tok/s398K ctx

NousResearch Hermes 4 14B

👁 Unsloth
Qwen3.5 9B

C48

9B12.2 GB99 tok/s558K ctx

dense

Qwen3.5 9B Uncensored HauhauCS Aggressive

starcoder2 15b instruct v0.1

SOLAR 10.7B Instruct v1.0 uncensored

starcoder2 15b i1

Qwen3.5 9B

Nous Hermes 2 SOLAR 10.7B

starcoder2 15b instruct v0.1

GGUF SOLARized GraniStral 14B 2102 YeAM HCT 32QKV

👁 UC Berkeley
Starling LM 7B

C47

7B11.9 GB98 tok/s8K ctx

dense

HelpingAI 15B i1

internlm JanusCoder 14B

GGUF SOLARized GraniStral 14B 1902 YeAM HCT

Yi Coder 9B Chat

glm 4 9b chat 1m

Meta Llama 3.1 8B Instruct

llava llama 3 8b v1 1

solar finalised finetuned Model 10.7B i1

👁 Unsloth
DeepSeek R1 0528 Qwen3 8B

C47

8B11.5 GB111 tok/s639K ctx

dense

Meta Llama 3 8B Instruct

Yi 1.5 9B Chat

👁 Unsloth
DeepSeek R1 Distill Llama 8B

C47

8B11.5 GB111 tok/s639K ctx

dense

SOLAR 10.7B v1.0

Dolphin3.0 Llama3.1 8B

Llama 3 8B Instruct 32k v0.1

Meta Llama 3.1 8B Instruct

vntl llama3 8b v2

DeepSeek R1 0528 Qwen3 8B

HelpingAI2.5 10B i1

10B13.0 GB89 tok/s494K ctx

DeepSeek R1 0528 Qwen3 8B

Qwen3 8B DeepSeek v3.2 Speciale Distill

Helply 10.2b chat i1

10.2B13.1 GB87 tok/s483K ctx

Hermes 3 Llama 3.1 8B

Llama 2 7B Chat

Yi 9B Coder i1

Mistral 7B Instruct v0.2

Hermes 2 Pro Llama 3 8B

Dolphin3.0 Llama3.1 8B

Mistral 7B Instruct v0.3

granite 8b code instruct 4k

HelpingAI2 9B

dolphin 2.9.4 llama3.1 8b

👁 Google
Gemma 3 1B

C47

1B6.7 GB14 tok/s33K ctx

dense

HelpingAI2 9B i1

openchat 3.6 8b 20240522 IMat

HelpingAI 9B 200k i1

aya expanse 8b

HelpingAI 9B i1

EXAONE 3.5 7.8B Instruct

dolphin v2 8b abliterated i1

EXAONE 3.5 7.8B Instruct

aya expanse 8b orthogonal heretic i1

EXAONE 3.5 7.8B Instruct i1

Mistral 7B Instruct v0.3

Hermes 2 Pro Mistral 7B

exaone 3.0 7.8b it

zephyr 7B beta

Nous Hermes 2 Mistral 7B DPO

aya expanse 8b orthogonal heretic

Falcon H1R 7B

falcon mamba 7b instruct Q4 K M

👁 Google
Gemma 2 2B

C47

2B8.5 GB28 tok/s8K ctx

dense

zephyr 7B alpha

speechless zephyr code functionary 7b

internlm2 5 1 8b chat i1

internlm3 8b instruct abliterated i1

StarCoder2 7B

japanese stablelm instruct gamma 7B

👁 Intel
Neural Chat 7B

C46

7B11.9 GB98 tok/s8K ctx

dense

Falcon H1 7B Instruct

👁 Unsloth
Falcon H1R 7B

C46

7B10.8 GB98 tok/s742K ctx

dense

👁 DeepSeek
DeepSeek LLM 67B

C46

67B52.4 GB9 tok/s4K ctx

dense

Mamba Codestral 7B v0.1

stabilityai japanese stablelm base gamma 7b

baichuan2 7b chat

DiscoPOP zephyr 7b gemma

starcoder2 7b

zephyr 7b beta Mistral 7B Instruct v0.2

OpenChat 3.5 7B Qwen v2.0 i1

OpenChat 3.5 7B Starling v2.0 i1

internlm2 math plus 7b IMat

CodeNinja 1.0 OpenChat 7B i1

internlm2 5 7b chat i1

Yi 1.5 6B Chat

OpenSafetyLab MD Judge v0 2 internlm2 7b

MD Judge v0 2 internlm2 7b i1

zephyr 7b gemma sft african ultrachat 100k

jointpreferences mistral 7b sft helpful

zephyr 7b dpo full i1

blossom v3 baichuan2 7b i1

blossom v1 baichuan 7b i1

stablelm 2 zephyr 1 6b

👁 01.AI
Yi 1.5 6B

C46

6B10.3 GB84 tok/s4K ctx

dense

Yi 1.5 6B Chat

stablelm 2 1 6b chat imatrix

👁 BigCode
StarCoder2 7B

C46

7B10.5 GB98 tok/s16K ctx

dense

HelpingAI2 6B

HelpingAI2 6B i1

5B9.3 GB70 tok/s1072K ctx

👁 NVIDIA
Nemotron Mini 4B

C46

4B10.1 GB56 tok/s4K ctx

dense

👁 Alibaba
Qwen 2.5 Coder 0.5B

C45

0.5B6.2 GB7 tok/s131K ctx

dense

HelpingAI2.5 5B i1

C45

4B8.6 GB56 tok/s1361K ctx

👁 Unsloth
Qwen3.5 4B

C44

4B8.6 GB56 tok/s1361K ctx

dense

gemma 3 4b it

C44

4B8.6 GB56 tok/s1361K ctx

gemma 3 4b it

C44

👁 Alibaba
Qwen 3 0.6B

C44

0.6B6.9 GB8 tok/s33K ctx

dense

👁 Meta
Llama 3.2 1B

C44

1B6.8 GB14 tok/s128K ctx

dense

Llama 3.2 3B Instruct

3B8.2 GB42 tok/s1827K ctx

👁 Qwen
Qwen2.5 3B Instruct

C43

3B7.9 GB42 tok/s1842K ctx

dense

Llama 3.2 3B Instruct

Ministral 3 3B Instruct 2512

HELVETE 3B

Hermes 3 Llama 3.2 3B

stablelm zephyr 3b

StarCoder2 3B

AI21 Jamba Reasoning 3B

stablelm 3b 4e1t

ai21labs AI21 Jamba Reasoning 3B

ai21labs AI21 Jamba2 3B

HelpingAI 3B hindi i1

AI21 Jamba2 3B

HelpingAI 3B hindi

AI21 Jamba2 3B i1

gemma 2 2b it

2B7.6 GB28 tok/s2776K ctx

EXAONE 3.5 2.4B Instruct

2.4B7.4 GB34 tok/s2323K ctx

2B7.2 GB28 tok/s2804K ctx

👁 Google
gemma 2b

C42

2B7.2 GB28 tok/s2804K ctx

dense

Gemmasutra Mini 2B v1

C41

2B7.2 GB28 tok/s2804K ctx

👁 BigCode
StarCoder2 3B

C41

3B8.0 GB42 tok/s16K ctx

dense

gemma 2 2b it

C41

1.5B6.8 GB21 tok/s3767K ctx

👁 Alibaba
Qwen 3.5 0.6B

C41

0.6B6.9 GB8 tok/s131K ctx

dense

👁 Qwen
Qwen2.5 1.5B Instruct

C40

1.5B6.8 GB21 tok/s3767K ctx

dense

👁 Unsloth
DeepSeek R1 Distill Qwen 1.5B

C40

1.5B6.8 GB21 tok/s3767K ctx

dense

Yi Coder 1.5B Chat

C40

1.6B6.9 GB22 tok/s3526K ctx

stablelm 2 zephyr 1.6b

C40

logos16v2 stablelm2 1.6b i1

C40

1.6B6.9 GB22 tok/s3526K ctx

1.5B6.8 GB21 tok/s3767K ctx

F Tier -- Can't Run (56 models)

Too heavy for this hardware.

Yi Coder 1.5B

D40

👁 Unsloth
Falcon H1 1.5B Instruct

D40

1.5B6.8 GB21 tok/s3767K ctx

dense

TinyLlama 1.1B Chat v1.0

1.1B6.5 GB15 tok/s5167K ctx

EXAONE 4.0 1.2B

1.2B6.6 GB17 tok/s4730K ctx

Llama 3.2 1B Instruct Q8 0

👁 Alibaba
Qwen 2.5 0.5B

D39

0.5B6.2 GB7 tok/s131K ctx

dense

1B6.6 GB14 tok/s5663K ctx

TinyLlama 1.1B Chat v0.3

1.1B6.5 GB15 tok/s5167K ctx

👁 TinyLlama
TinyLlama 1.1B Chat v0.6

D39

1.1B6.5 GB15 tok/s5167K ctx

dense

Llama 3.2 1B Instruct

1B6.4 GB14 tok/s5692K ctx

gemma 3 1b it

1B6.4 GB14 tok/s5692K ctx

TinyLlama 1.1B Chat v1.0 imatrix

1.1B6.5 GB15 tok/s5167K ctx

Falcon3 1B Instruct abliterated

D38

1B6.4 GB14 tok/s5692K ctx

0.5B6.2 GB7 tok/s6702K ctx

SmolVLM 500M Instruct

D37

Solar Open 69B REAP i1

D36

69B55.9 GB7 tok/s4K ctx

0.3B6.0 GB4 tok/s6729K ctx

embeddinggemma 300M

D35

granite embedding 107m multilingual

D33

0.11B5.9 GB2 tok/s6758K ctx

Falcon H1 Tiny 90M Instruct

D33

0.09B5.9 GB2 tok/s6759K ctx

Llama 3.3 70B Instruct

D30

70B56.6 GB7 tok/s4K ctx

stabilityai japanese stablelm instruct beta 70b

D29

70B56.6 GB7 tok/s4K ctx

230B149.8 GB2 tok/s4K ctx

👁 Alibaba
Qwen 3 235B A22B

235B151.9 GB2 tok/s4K ctx

moe

👁 Alibaba
Qwen3-Coder 480B A35B Instruct

480B301.4 GB2 tok/s4K ctx

moe

👁 Alibaba
Qwen 3.5 122B A10B

122B82.6 GB5 tok/s4K ctx

moe

👁 Alibaba
Qwen 3.5 397B A17B

397B250.7 GB2 tok/s4K ctx

moe

👁 Mistral
Mixtral 8x22B

141B95.1 GB2 tok/s4K ctx

moe

👁 Mistral
Devstral 2 123B Instruct

123B86.1 GB2 tok/s4K ctx

dense

👁 Mistral
Mistral Small 4 119B

119B83.7 GB6 tok/s4K ctx

moe

👁 Mistral
Leanstral 119B A6B

119B87.1 GB5 tok/s4K ctx

moe

👁 Mistral
Mistral Large 3

675B422.8 GB2 tok/s4K ctx

+1moe

👁 Z.ai
GLM-5

744B478.6 GB2 tok/s4K ctx

moe

👁 Z.ai
GLM-5.1

754B484.7 GB2 tok/s4K ctx

moe

👁 Moonshot AI
Kimi K2.5

1000B623.1 GB2 tok/s4K ctx

moe

👁 Moonshot AI
Kimi K2.6

1000B623.1 GB2 tok/s4K ctx

+1moe

👁 DeepSeek
DeepSeek V3 671B

671B474.6 GB2 tok/s4K ctx

moe

👁 DeepSeek
DeepSeek V3.1 671B

671B474.6 GB2 tok/s4K ctx

moe

👁 DeepSeek
DeepSeek V4 Pro

1600B869.6 GB2 tok/s4K ctx

moe

👁 DeepSeek
DeepSeek V4 Flash

284B165.0 GB2 tok/s4K ctx

moe

👁 DeepSeek
DeepSeek R1 671B

671B474.6 GB2 tok/s4K ctx

moe

👁 DeepSeek
DeepSeek Coder V2 236B

236B208.3 GB2 tok/s4K ctx

moe

👁 DeepSeek
DeepSeek V2.5 236B

236B208.3 GB2 tok/s4K ctx

moe

👁 Meta
Llama 4 Scout 17B 16E

109B75.1 GB6 tok/s4K ctx

moe

👁 Meta
Llama 4 Maverick 17B 128E

400B252.6 GB2 tok/s4K ctx

moe

👁 Cohere
Command R+ 104B

104B72.6 GB3 tok/s4K ctx

dense

👁 Cohere
Command A 111B

111B77.3 GB2 tok/s4K ctx

dense

👁 Mistral AI
Pixtral Large 124B

124B86.7 GB2 tok/s4K ctx

dense

👁 Meta
Llama 3.1 405B

405B260.4 GB2 tok/s4K ctx

dense

👁 OpenAI
GPT-OSS 120B

117B82.0 GB2 tok/s4K ctx

dense

👁 DeepSeek
DeepSeek V3.2

671B415.5 GB2 tok/s4K ctx

moe

MiniMax M2.7

moe

Solar Open 100B

100B78.4 GB2 tok/s4K ctx

236B177.3 GB2 tok/s4K ctx

K EXAONE 236B A23B

235B176.6 GB2 tok/s4K ctx

Baichuan M3 235B

235B176.6 GB2 tok/s4K ctx

Baichuan M3 235B i1

BaichuanMed OCR 72B i1

72B58.1 GB6 tok/s4K ctx

Solar Open 100B i1

100B78.4 GB2 tok/s4K ctx

👁 Unsloth
Qwen3.5 122B A10B

122B79.8 GB2 tok/s4K ctx

dense

👁 Unsloth
Qwen3.5 397B A17B

397B294.4 GB2 tok/s4K ctx

dense

Compare with...

See How Other Hardware Stacks Up

Check tier lists for similar GPUs and Apple Silicon configs

RTX 6000 Ada 48GB

48 GB VRAM

RTX PRO 5000 Blackwell 48GB

48 GB VRAM

NVIDIA L40S 48GB

48 GB VRAM

Radeon Pro W7900 48GB

48 GB VRAM