Qualcomm · Snapdragon X Elite (X1E-80-100) · Laptop

Best AI Models for Snapdragon X Elite Copilot+ PC (32 GB)

Runs 35B MoE models well; 9B, 14B and 27–31B fit but are slow (14, 9 and 5 tok/s), and 70B doesn't fit in 26 GB.

32 GB unified26 GB for models135 GB/s9 TFLOPS FP162024

What LLMs can the Snapdragon X Elite Copilot+ PC run?

#	Model	Speed	Quant · Memory
1	Qwen3.6 35B A3BAlibabaHow to run → Q4_K_M · 22.0 GB · 27.0 t/s	27.0 tok/s	Q4_K_M · 22.0 GB
2	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 29.1 t/s	29.1 tok/s	Q4_K_M · 16.5 GB
3	Qwen3 30B A3B Thinking 2507Alibaba Q4_K_M · 19.4 GB · 35.4 t/s	35.4 tok/s	Q4_K_M · 19.4 GB
4	GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 31.3 t/s	31.3 tok/s	Q4_K_M · 13.1 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 46.1 t/s	46.1 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · Memory
1	Qwen3.6 35B A3BAlibabaHow to run → Q4_K_M · 22.0 GB · 27.0 t/s	27.0 tok/s	Q4_K_M · 22.0 GB
2	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 29.1 t/s	29.1 tok/s	Q4_K_M · 16.5 GB
3	GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 31.3 t/s	31.3 tok/s	Q4_K_M · 13.1 GB
4	Llama 3.1 8B InstructMeta Q4_K_M · 7.9 GB · 16.6 t/s	16.6 tok/s	Q4_K_M · 7.9 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 46.1 t/s	46.1 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · Memory
1	Gemma 4 26B A4B ITGoogleHow to run → Q4_K_M · 16.5 GB · 29.1 t/s	29.1 tok/s	Q4_K_M · 16.5 GB
2	Qwen3 30B A3B Thinking 2507Alibaba Q4_K_M · 19.4 GB · 35.4 t/s	35.4 tok/s	Q4_K_M · 19.4 GB
3	Qwen3.5 4BAlibaba Q4_K_M · 3.4 GB · 26.9 t/s	26.9 tok/s	Q4_K_M · 3.4 GB
4	GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 31.3 t/s	31.3 tok/s	Q4_K_M · 13.1 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 46.1 t/s	46.1 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed ↓	Quant · Memory
1	Qwen3 30B A3B Thinking 2507AlibabaHow to run → Q4_K_M · 19.4 GB · 35.4 t/s	35.4 tok/s	Q4_K_M · 19.4 GB
2	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 46.1 t/s	46.1 tok/s	Q4_K_M · 19.3 GB
3	Qwen3 30B A3B Instruct 2507Alibaba Q4_K_M · 19.4 GB · 35.4 t/s	35.4 tok/s	Q4_K_M · 19.4 GB
4	GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 31.3 t/s	31.3 tok/s	Q4_K_M · 13.1 GB
5	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 29.1 t/s	29.1 tok/s	Q4_K_M · 16.5 GB

#	Model	Speed	Quant · Memory
1	Qwen3.6 35B A3BAlibabaHow to run → Q4_K_M · 22.0 GB · 27.0 t/s	27.0 tok/s	Q4_K_M · 22.0 GB
2	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 29.1 t/s	29.1 tok/s	Q4_K_M · 16.5 GB
3	Qwen3.5 4BAlibaba Q4_K_M · 3.4 GB · 26.9 t/s	26.9 tok/s	Q4_K_M · 3.4 GB
4	Gemma 3 4B ITGoogle Q4_K_M · 5.0 GB · 30.9 t/s	30.9 tok/s	Q4_K_M · 5.0 GB
5	MiniCPM V 4.7 35B A3BOpenBMB Q4_K_M · 22.1 GB · 31.4 t/s	31.4 tok/s	Q4_K_M · 22.1 GB

See all 337 models on the Snapdragon X Elite Copilot+ PC →Bars = traction for models without a Score yet

About the Snapdragon X Elite Copilot+ PC

The Snapdragon X Elite Copilot+ PC is a 32 GB laptop with more memory than speed: it runs 35B MoE models well. The system keeps about 6 GB, leaving 26 GB for models. Prompt reading is its weak spot: a 4K-token prompt waits 14 s before the first word on Qwen3.6 35B A3B. Snapdragon X Elite (X1E-80-100), released 2024.

Speed in 32 GB#8135 GB/s

Prompt reading#99 TFLOPS

For models26 GBThe system keeps 6 GB

ChipSnapdragon X Elite (X1E-80-100)released 2024

Snapdragon X Elite Copilot+ PC strengths and weaknesses

Strengths

Runs 35B MoE models: Qwen3.6 35B A3B at 27.0 tok/s in 22 GB

Weaknesses

70B models don't fit: Llama 3.3 70B Instruct needs 51.3 GB at Q4 vs 26 GB usable here
Slow on large dense models: Qwen3.8 27B at 5.0 tok/s
Slow prompt reading: A 32K-token prompt on Qwen3.6 35B A3B waits 1.9 min for the first token

Memory for models

26 GB of 32 GB

9B6.6 GB14B11.9 GB27–31B17.9 GB35B MoE22 GB

9B6.6 GB14B11.9 GB35B MoE22 GB

For models 26 GBThe system keeps 6 GBModel size at Q4

70B models need 51.3 GB at Q4, more than the 26 GB available.Or rent an A100 (80 GB) from $1.31/hr →

Similar devices and alternatives

The same chip from other makers first, then machines and GPUs people weigh against it, each with the one reason it is worth a look. Speeds on Qwen3.6 35B A3B.

Newer machine, 13% slowerMac Mini M4 (32 GB)Apple · 28.5 GB for models · M4

▼ 13% speed▲ 2.5 GB$999

Newer laptop, 6% fasterMacBook Pro 14 M5Apple · 28.5 GB for models · M5

▲ 6% speed▲ 2.5 GB$1,839

104% faster, 6.5 GB more memoryMac Studio M4 Max (36 GB)Apple · 32.5 GB for models · M4 Max

▲ 104% speed▲ 6.5 GB$2,499

Newer machine, 72% fasterMac Mini M4 Pro (24 GB)Apple · 20.5 GB for models · M4 Pro

▲ 72% speed▼ 5.5 GB$1,599

Reads prompts 27× faster, 24 GB for modelsRTX 4090NVIDIA · 24 GB VRAM · GPU

▲ 27× prompts▲ 459% speed▼ 2 GB$1,599rent $0.34/hr

Reads prompts 34× faster, 32 GB for modelsRTX 5090NVIDIA · 32 GB VRAM · GPU

▲ 34× prompts▲ 690% speed▲ 6 GB$5,376rent $0.50/hr

Snapdragon X Elite Copilot+ PC tokens per second by model size

Output speed depends on memory bandwidth (135 GB/s). Grey rows need more than the 26 GB available.

9BQwen3.5 9B · Q4_K_M · 6.6 GB
13.8tok/s
14BQwen3 14B · Q4_K_M · 11.9 GB slow
9.2tok/s
27–31BQwen3.8 27B · Q4_K_M · 17.9 GB slow
5.0tok/s
35B MoEQwen3.6 35B A3B · Q4_K_M · 22.0 GB
27.0tok/s
70BLlama 3.3 70B Instruct · not at 4-bit
needs 51.3 GB
120B MoEGPT OSS 120B · not at 4-bit
needs 70.8 GB

Estimated speeds: Q4_K_M, 4K context, one user. How we estimate

Unified-memory platforms: writing vs compute

Writing, tok/sCompute, TFLOPS

M224 GB · 100 GB/s
22.1
7
Orin NX16 GB · 102.4 GB/s
37.0
15
M324 GB · 102.4 GB/s
22.6
7
Ryzen AI 9 HX 37032 GB · 120 GB/s
20.7
12
M432 GB · 120 GB/s
23.5
9
Snapdragon X Elite (X1E-80-100)32 GB · 135 GB/s
27.0
9

Machines with about the same memory bandwidth write at about the same speed; compute decides how fast prompts are read. Speeds on Qwen3.6 35B A3B.

All 1158 models on the Snapdragon X Elite Copilot+ PC

Featured only

Best quant that fits in 26 GB, ordered by llmrun Score. Fit grades how well it runs here, S best. How ranking works

LLM models compatible with Snapdragon X Elite Copilot+ PC — ranked by Score, recency and Fit
Model	Quant	VRAM	Speed	Context	Fit
XiYanSQL QwenCoder 32B 250432BChatCode Q4_K_M·4.4 t/s	Q4_K_M	23.8 GB92%	4.4 t/s	33K	D23, Barely runs
Starcoder2 15B16.0BChatCode Q4_K_M·8.7 t/s	Q4_K_M	11.2 GB43%	8.7 t/s	16K	B46, Decent
Yi 6B6.1BChat Q4_K_M·21.6 t/s	Q4_K_M	4.2 GB16%	21.6 t/s	4K	B62, Decent
Qwen 1 8B1.8BChat Q4_K_M·72.5 t/s	Q4_K_M	1.9 GB7%	72.5 t/s	8K	S86, Runs great
Qwen2 0.5B494MChat Q4_K_M·141.5 t/s	Q4_K_M	0.8 GB3%	141.5 t/s	131K	S96, Runs great
Granite 3.1 8B Instruct8.2BChatFunctions Q4_K_M·15.8 t/s	Q4_K_M	7.9 GB30%	15.8 t/s	131K	B55, Decent
Starcoder15.8BChatCode Q4_K_M·8.4 t/s	Q4_K_M	13.6 GB52%	8.4 t/s	—	B46, Decent
Qwen2.5 Math 1.5B1.5BChatMathFunctions Q4_K_M·68.6 t/s	Q4_K_M	1.3 GB5%	68.6 t/s	4K	S85, Runs great
Llama Guard 3 8B8.0BChat BF16·5.0 t/s	BF16	20.2 GB78%	5.0 t/s	—	C42, Usable
Qwen2 7B Instruct7.6BChat Q4_K_M·17.6 t/s	Q4_K_M	5.8 GB22%	17.6 t/s	33K	B57, Decent
Phi 3 Mini 128k Instruct3.8BChatCode Q4_K_M·25.8 t/s	Q4_K_M	9.0 GB35%	25.8 t/s	131K	A66, Runs well
Shieldgemma 2B2.6BChat Q4_K_M·50.7 t/s	Q4_K_M	3.6 GB14%	50.7 t/s	—	A80, Runs well
Nemotron H 8B Reasoning 128K8.1BChatReasoning BF16·4.9 t/s	BF16	20.4 GB78%	4.9 t/s	—	C42, Usable
Gemma 1.1 2B IT2.5BChat Q4_K_M·53.2 t/s	Q4_K_M	3.5 GB13%	53.2 t/s	—	A81, Runs well
Qwen1.5 1.8B Chat1.8BChat Q4_K_M·48.8 t/s	Q4_K_M	4.6 GB18%	48.8 t/s	33K	A79, Runs well

Snapdragon X Elite Copilot+ PC specs

Specs

Chip: Snapdragon X Elite (X1E-80-100) · Qualcomm Adreno X1
CPU cores: 12
Unified memory: 32 GB LPDDR5X-8448
For models: 26 GB (The system keeps 6 GB)
Memory bandwidth: 135 GB/s
FP16 (dense): 9.2 TFLOPS
NPU: Hexagon NPU (45 TOPS)
Released: 2024-06-18

Snapdragon X Elite Copilot+ PC FAQ

How much memory is available for AI on the Snapdragon X Elite Copilot+ PC?

About 26 GB of the 32 GB. The system keeps roughly 6 GB for itself and other apps.

What is the largest model the Snapdragon X Elite Copilot+ PC can run?

Qwen3.6 35B A3B (36.0B) is the largest at 4-bit or better: at Q4_K_M it uses 22.0 GB, 85% of the available memory, and runs at about 27.0 tok/s.

Can the Snapdragon X Elite Copilot+ PC run 70B models?

No. Llama 3.3 70B Instruct needs 51.3 GB at Q4, more than the 26 GB available.

How long does the Snapdragon X Elite Copilot+ PC take to read a long prompt?

On Qwen3.6 35B A3B, about 14 s for a 4K-token prompt and 1.9 min for a 32K-token document before the first word appears.

Is renting a GPU cheaper than buying the Snapdragon X Elite Copilot+ PC?

An RTX 5090 (32 GB) rents from $0.45/hr.

24 GB GPUs Laptops Mac Studio M1 Max Mac Studio M2 Max MacBook Pro 14" M1 Max MacBook Pro 14" M2 Max MacBook Pro 16" M1 Max MacBook Pro 16" M2 Max Snapdragon X2 Elite Extreme Copilot+ PC Best coding LLMs All hardware