Qualcomm · Snapdragon X2 Elite Extreme (X2E-96-100) · Laptop

Best AI Models for Snapdragon X2 Elite Extreme Copilot+ PC (48 GB)

Runs 14B and 35B MoE models well; dense 27–31B fits but is slow at 9 tok/s, and 70B only just fits in 44 GB.

48 GB unified44 GB for models228 GB/s15 TFLOPS FP162026

What LLMs can the Snapdragon X2 Elite Extreme Copilot+ PC run?

#	Model	Speed	Quant · Memory
1	Qwen3.6 35B A3BAlibabaHow to run → Q4_K_M · 22.0 GB · 42.5 t/s	42.5 tok/s	Q4_K_M · 22.0 GB
2	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 46.4 t/s	46.4 tok/s	Q4_K_M · 16.5 GB
3	Qwen3 30B A3B Thinking 2507Alibaba Q4_K_M · 19.4 GB · 53.6 t/s	53.6 tok/s	Q4_K_M · 19.4 GB
4	GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 50.3 t/s	50.3 tok/s	Q4_K_M · 13.1 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 66.9 t/s	66.9 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 · 42.5 t/s	42.5 tok/s	Q4_K_M · 22.0 GB
2	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 46.4 t/s	46.4 tok/s	Q4_K_M · 16.5 GB
3	GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 50.3 t/s	50.3 tok/s	Q4_K_M · 13.1 GB
4	Llama 3.1 8B InstructMeta Q4_K_M · 7.9 GB · 28.0 t/s	28.0 tok/s	Q4_K_M · 7.9 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 66.9 t/s	66.9 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 · 42.5 t/s	42.5 tok/s	Q4_K_M · 22.0 GB
2	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 46.4 t/s	46.4 tok/s	Q4_K_M · 16.5 GB
3	Qwen3.5 9BAlibaba Q4_K_M · 6.6 GB · 23.3 t/s	23.3 tok/s	Q4_K_M · 6.6 GB
4	GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 50.3 t/s	50.3 tok/s	Q4_K_M · 13.1 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 66.9 t/s	66.9 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 · 53.6 t/s	53.6 tok/s	Q4_K_M · 19.4 GB
2	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 66.9 t/s	66.9 tok/s	Q4_K_M · 19.3 GB
3	Qwen3 30B A3B Instruct 2507Alibaba Q4_K_M · 19.4 GB · 53.6 t/s	53.6 tok/s	Q4_K_M · 19.4 GB
4	GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 50.3 t/s	50.3 tok/s	Q4_K_M · 13.1 GB
5	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 46.4 t/s	46.4 tok/s	Q4_K_M · 16.5 GB

See all 367 models on the Snapdragon X2 Elite Extreme Copilot+ PC →Bars = traction for models without a Score yet

About the Snapdragon X2 Elite Extreme Copilot+ PC

The Snapdragon X2 Elite Extreme Copilot+ PC is a 48 GB laptop with more memory than speed: it runs 14B and 35B MoE models well. The system keeps about 4 GB, leaving 44 GB for models. Prompt reading is its weak spot: a 4K-token prompt waits 8.5 s before the first word on Qwen3.6 35B A3B. Snapdragon X2 Elite Extreme (X2E-96-100), released 2026.

Speed in 48 GB#6228 GB/s

Prompt reading#615 TFLOPS

For models44 GBThe system keeps 4 GB

ChipSnapdragon X2 Elite Extreme (X2E-96-100)released 2026

Snapdragon X2 Elite Extreme Copilot+ PC strengths and weaknesses

Strengths

Runs 35B MoE models: Qwen3.6 35B A3B at 42.5 tok/s in 22 GB

Weaknesses

120B MoE models don't fit: GPT OSS 120B needs 70.8 GB at Q4 vs 44 GB usable here
Slow on large dense models: Llama 3.3 70B Instruct at 3.8 tok/s, Qwen3.8 27B at 8.5 tok/s
Slow prompt reading: A 32K-token prompt on Qwen3.6 35B A3B waits 1.1 min for the first token

Memory for models

44 GB of 48 GB

9B6.6 GB14B11.9 GB27–31B17.9 GB35B MoE22 GB70B43.5 GB

35B MoE22 GB70B43.5 GB

For models 44 GBThe system keeps 4 GBModel size at Q4

120B MoE models need 70.8 GB at Q4, more than the 44 GB available.Or rent an RTX PRO 6000 Blackwell Max-Q Workstation Edition (96 GB) from $1.14/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.

51% faster, same memoryMacBook Pro 16" M4 Max (48 GB)Apple · 44.5 GB for models · M4 Max

▲ 51% speedsame memory$3,999

Newer machine, as fastMac Mini M4 Pro (48 GB)Apple · 44.5 GB for models · M4 Pro

same speedsame memory$2,199

9% faster, 16.5 GB more memoryMacBook Pro 14 M5 ProApple · 60.5 GB for models · M5 Pro

▲ 9% speed▲ 16.5 GB$2,299

Newer machine, 45% slowerMac Mini M4 (32 GB)Apple · 28.5 GB for models · M4

▼ 45% speed▼ 15.5 GB$999

3.6× faster if your models fit in 24 GBRTX 4090NVIDIA · 24 GB VRAM · GPU

▲ 255% speed▼ 20 GB$1,599rent $0.34/hr

GPU, 402% fasterRTX 5090NVIDIA · 32 GB VRAM · GPU

▲ 402% speed▼ 12 GB$5,376rent $0.48/hr

Snapdragon X2 Elite Extreme Copilot+ PC tokens per second by model size

Output speed depends on memory bandwidth (228 GB/s). Grey rows need more than the 44 GB available.

9BQwen3.5 9B · Q4_K_M · 6.6 GB
23.3tok/s
14BQwen3 14B · Q4_K_M · 11.9 GB
15.6tok/s
27–31BQwen3.8 27B · Q4_K_M · 17.9 GB slow
8.5tok/s
35B MoEQwen3.6 35B A3B · Q4_K_M · 22.0 GB
42.5tok/s
70BLlama 3.3 70B Instruct · Q4_0 · 43.5 GB tight · Q4_0
3.8tok/s
120B MoEGPT OSS 120B · not at 4-bit
needs 70.8 GB

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

All 1215 models on the Snapdragon X2 Elite Extreme Copilot+ PC

Featured only

Best quant that fits in 44 GB, ordered by llmrun Score. Fit grades how well it runs here, S best. How ranking works

LLM models compatible with Snapdragon X2 Elite Extreme Copilot+ PC — ranked by Score, recency and Fit
Model	Quant	VRAM	Speed	Context	Fit
Qwen3 1.7B2.0BChatFunctions Q4_K_M·84.7 t/s	Q4_K_M	3.4 GB8%	84.7 t/s	41K	S88, Runs great
Magistral Small 250623.6BChat Q4_K_M·10.0 t/s	Q4_K_M	17.8 GB40%	10.0 t/s	41K	B48, Decent
Mistral Medium 3.5 128B127.7BChatFunctions IQ2_XXS·4.1 t/s	IQ2_XXS	41.3 GB94%	4.1 t/s	262K	D19, Barely runs
Riva Translate 4B Instruct4.2BChat Q4_K_M·49.1 t/s	Q4_K_M	3.7 GB8%	49.1 t/s	8K	A79, Runs well
Qwen3.5 35B A3B Base36.0BVision Q4_K_M·42.5 t/s	Q4_K_M	22.0 GB50%	42.5 t/s	262K	A76, Runs well
Llama Poro 2 8B Instruct8.0BChat Q4_K_M·27.5 t/s	Q4_K_M	6.2 GB14%	27.5 t/s	8K	A68, Runs well
GPT Neo 125M150MChat BF16·449.1 t/s	BF16	0.3 GB1%	449.1 t/s	2K	S100, Runs great
Salamandra 2B Instruct2.3BChat Q4_K_M·72.3 t/s	Q4_K_M	3.3 GB7%	72.3 t/s	8K	S86, Runs great
Japanese GPT Neox Small204MChat Q4_K_M·1140.0 t/s	Q4_K_M	0.1 GB0%	1140.0 t/s	2K	S100, Runs great
Gemma 3 4B IT4.3BVision Q4_K_M·52.2 t/s	Q4_K_M	5 GB11%	52.2 t/s	—	A81, Runs well
Param 1 5B5BChat BF16·14.2 t/s	BF16	10.6 GB24%	14.2 t/s	4K	B53, Decent
Jan v3 4B Base Instruct4.4BChatCode Q4_K_M·47.2 t/s	Q4_K_M	4.5 GB10%	47.2 t/s	262K	A78, Runs well
Bloom 560M559MChat Q4_K_M·400.5 t/s	Q4_K_M	1.6 GB4%	400.5 t/s	—	S100, Runs great
Phi 4 Reasoning Plus14.7BChatMathCodeReasoning Q4_K_M·15.6 t/s	Q4_K_M	12.5 GB28%	15.6 t/s	33K	B55, Decent
SmolLM2 135M Instruct135MChat Q4_K_M·344.7 t/s	Q4_K_M	0.6 GB1%	344.7 t/s	8K	S100, Runs great

Snapdragon X2 Elite Extreme Copilot+ PC specs

Specs

Chip: Snapdragon X2 Elite Extreme (X2E-96-100) · Qualcomm Adreno X2-90
CPU cores: 18
Unified memory: 48 GB LPDDR5X-9523
For models: 44 GB (The system keeps 4 GB)
Memory bandwidth: 228 GB/s
FP16 (dense): 15.2 TFLOPS
NPU: Hexagon NPU (80 TOPS INT8)
Released: 2026-01-06

Snapdragon X2 Elite Extreme Copilot+ PC FAQ

How much memory is available for AI on the Snapdragon X2 Elite Extreme Copilot+ PC?

About 44 GB of the 48 GB. The system keeps roughly 4 GB for itself and other apps.

What is the largest model the Snapdragon X2 Elite Extreme Copilot+ PC can run?

Llama 3.3 70B Instruct (70.6B) is the largest at 4-bit or better: at Q4_0 it uses 43.5 GB, 99% of the available memory, and runs at about 3.8 tok/s.

Can the Snapdragon X2 Elite Extreme Copilot+ PC run Qwen3.6 35B A3B?

Yes. Qwen3.6 35B A3B fits at Q4_K_M, using 22.0 GB of the 44 GB available, and it is the best model that runs well on this machine.

Can the Snapdragon X2 Elite Extreme Copilot+ PC run 120B MoE models?

No. GPT OSS 120B needs 70.8 GB at Q4, more than the 44 GB available.

How long does the Snapdragon X2 Elite Extreme Copilot+ PC take to read a long prompt?

On Qwen3.6 35B A3B, about 8.5 s for a 4K-token prompt and 68 s for a 32K-token document before the first word appears.

Is renting a GPU cheaper than buying the Snapdragon X2 Elite Extreme Copilot+ PC?

An A100 (80 GB) rents from $1.00/hr.

24 GB GPUs Laptops Mac Studio M5 Max MacBook Pro 14" M5 Max MacBook Pro 16" M5 Max Mac Studio M4 Max MacBook Pro 14" M4 Max Best coding LLMs All hardware