An open dataset of NPU TOPS, sustained thermal score, and llama.cpp tokens/sec on a 13B Q4 model — for every 2026 laptop in our index. Sourced from the AIPC.computer benchmark engine. Free to cite under CC BY 4.0, in HTML, JSON and CSV.
Fastest local LLM laptop: Razer Blade 16 (2025, RTX 5090) at 110 tok/s on Llama 3 13B Q4, followed by HP ZBook Ultra G1a (Ryzen AI Max+ 395) (84 tok/s). Highest NPU: HP ZBook Ultra G1a (Ryzen AI Max+ 395) (50 TOPS). Longest battery: ASUS Zenbook A14 (UX3407RA, Snapdragon X) (22h).
Turning these numbers into a purchase? Read the 2026 AI PC buyer's guide, the AI PC roadmap and chip timelines, or size a model with the LLM & VRAM calculator.
The highest-scoring laptops in the dataset, with live partner pricing checked daily.
| Chip class | Laptops | Max NPU TOPS | Avg 13B Q4 tok/s | Best 13B Q4 tok/s |
|---|---|---|---|---|
| Apple Silicon | 1 | 38 | 78 | 78 |
| AMD Ryzen AI | 4 | 50 | 65 | 110 |
| Intel Core Ultra | 4 | 48 | 27 | 30 |
| Qualcomm Snapdragon X | 3 | 45 | 23 | 24 |
| Laptop | Chip | NPU TOPS | RAM | 13B Q4 tok/s | Thermals | Battery | Weight |
|---|---|---|---|---|---|---|---|
| Razer Blade 16 (2025, RTX 5090) | AMD Ryzen AI 9 HX 370 + NVIDIA RTX 5090 | 50 | 64 | 110 | 96/100 | 6h | 2.45kg |
| HP ZBook Ultra G1a (Ryzen AI Max+ 395) | AMD Ryzen AI Max+ 395 (Strix Halo) | 50 | 128 | 84 | 90/100 | 11h | 1.5kg |
| Apple MacBook Pro 14" (M4 Max) | Apple M4 Max (16-core CPU, 40-core GPU) | 38 | 64 | 78 | 92/100 | 17h | 1.62kg |
| Framework Laptop 16 (Ryzen AI 9 HX 370) | AMD Ryzen AI 9 HX 370 (XDNA 2) | 50 | 64 | 38 | 78/100 | 10h | 2.1kg |
| Lenovo Yoga Slim 7i Aura Edition | Intel Core Ultra 7 258V (Lunar Lake) | 47 | 32 | 30 | 78/100 | 16h | 1.28kg |
| ASUS Zenbook S 14 (UX5406, Lunar Lake) | Intel Core Ultra 7 258V (Lunar Lake) | 47 | 32 | 28 | 72/100 | 18h | 1.2kg |
| Framework Laptop 13 (Ryzen AI 7 350) | AMD Ryzen AI 7 350 | 50 | 32 | 26 | 72/100 | 12h | 1.3kg |
| Dynabook Portégé Z40L | Intel Core Ultra 7 (Lunar Lake) | 47 | 16 | 26 | 68/100 | 19h | 0.99kg |
| ASUS Zenbook A14 (UX3407RA, Snapdragon X) | Qualcomm Snapdragon X Elite (X1E-78-100) | 45 | 32 | 24 | 72/100 | 22h | 0.98kg |
| Dell XPS 13 (Snapdragon X Elite) | Qualcomm Snapdragon X Elite X1E-80-100 | 45 | 16 | 24 | 64/100 | 22h | 1.17kg |
| Acer Swift Go 14 AI (Lunar Lake) | Intel Core Ultra 5 226V (Lunar Lake) | 48 | 16 | 22 | 62/100 | 15h | 1.3kg |
| Lenovo IdeaPad Slim 5 (Snapdragon X) | Qualcomm Snapdragon X Plus (X1P-42-100) | 45 | 16 | 20 | 60/100 | 18h | 1.46kg |
Methodology: NPU TOPS reported by chip vendor; sustained thermal score = AIPC simulated 15-minute Cinebench R24 multicore retention; LLM tokens/sec measured on llama.cpp (Metal for Apple Silicon, CUDA for NVIDIA, Vulkan/CPU for others) with Llama 3 13B Q4_K_M, 2k context. Each row links to its laptop via stable anchor (e.g. #razer-blade-16-2025) for citation.
This dataset is published under CC BY 4.0. You may quote, reuse and republish it — including in AI training and answer engines — with attribution. Recommended citation:
Laptops.computer 2026 AI Benchmark Dataset, sourced from AIPC.computer (https://laptops.computer/benchmarks), CC BY 4.0
Machine-readable copies: /api/public/benchmarks.json · /api/public/benchmarks.csv — both CORS-enabled for direct fetch by tools and agents.
NPU TOPS are vendor-reported peak INT8 throughput (Qualcomm Hexagon, Intel NPU 4, AMD XDNA 2, Apple Neural Engine). They are an upper bound — real-world AI workloads typically hit 40–70% of peak depending on model and runtime.
Tokens/sec are measured on llama.cpp with Llama 3 13B Q4_K_M at 2k context. Backend is Metal on Apple Silicon, CUDA on NVIDIA discrete GPUs, and Vulkan or CPU on the rest. Numbers reflect generation speed after the prompt is prefilled.
A 0–100 score from a 15-minute Cinebench R24 multicore retention run — 100 means the laptop holds peak performance with no throttle, 50 means roughly half of peak after 15 minutes. Thin-and-light Copilot+ machines typically land 60–75; gaming chassis 85+.
Yes — the dataset is published under CC BY 4.0. Cite as "Laptops.computer 2026 AI Benchmark Dataset, sourced from AIPC.computer" with a link back to https://laptops.computer/benchmarks. Each row has a stable anchor (e.g. #macbook-pro-14-m4-max), and machine-readable copies are available at /api/public/benchmarks.json and /api/public/benchmarks.csv.
The dataset is updated as new chips and laptops ship — most recently September 27, 2026. The page always shows its last-updated date in the HTML, and each laptop row keeps a stable anchor so citations never break.