RTX laptops (5080/5090) for local AI: local AI test and review
Local AI that fits in a bag—16 or 24 GB of VRAM. Let's look at what matters for running LLMs locally : memory, bandwidth, speed, price—and who it's (or isn't) the right purchase for.
Updated on 09/10/2026
Where to buy it
The button configuration is the one offered for purchase. On mini PCs, keep some memory available for the system and check the inference engine; macOS/MLX and CUDA are not interchangeable.
Affiliate links — BestLLMfor may earn a commission from purchases at no extra cost to you, which does not influence these independently determined recommendations. As an Amazon Associate, BestLLMfor earns from qualifying purchases.
Technical specifications
- Memory16 GB (RTX 5080) or 24 GB (RTX 5090) dedicated GDDR7
- Bandwidth≈ 700–900 GB/s depending on the chassis TGP
- ComputeRTX 5080/5090 Laptop (Blackwell) + Intel Ultra 9 / Ryzen 9 CPU
- Power consumption100–175 W GPU (TGP varies by model)
- Indicative pricefrom ≈ €2,500 (RTX 5080 16 GB); from ≈ €4,300 (RTX 5090 24 GB)
- Largest model16 GB: ~14B comfortably in Q4 (24B is tight); 24 GB: 30B class in Q4 (Qwen3-32B, GLM-4.7-Flash) — 70B models remain out of reach; target a 128 GB mini-PC or a desktop for that
Who is it for?
Advantages and limitations
✓ Strengths
- Complete autonomy: your LLM follows you everywhere, even offline
- High GDDR7 bandwidth: excellent generation speed on models that fit
- Full CUDA support: same stack as on desktop (Ollama, llama.cpp, vLLM)
- Verified models in stock: Acer Predator Helios 18 AI (RTX 5080 16 GB + 64 GB RAM, ≈ €3,000), Gigabyte A18 Pro (RTX 5080, ≈ €2,500), MSI Vector 17 HX (RTX 5090 24 GB, ≈ €4,300), Acer Helios 18 AI Pro (RTX 5090 24 GB, ≈ €6,500)
✗ Limitations
- VRAM capped at 24 GB: no 70B, unlike 128 GB mini-PCs with unified memory
- TGP limits mobile RTX performance versus desktop (a 5090 laptop ≈ a 5080 desktop under sustained load)
- Heat and ventilation during long inference runs — plan for a ventilated stand
- Higher VRAM price per GB than the equivalent desktop
Verdict
Frequently asked questions
Can an RTX laptop (5080/5090) for local AI run a 70B LLM locally?
16 GB: approximately 14B at comfortable Q4 (24B is tight); 24 GB: 30B class at Q4 (Qwen3-32B, GLM-4.7-Flash)—70B models remain out of reach; get a 128 GB mini-PC or a desktop for those.
How much does an RTX laptop (5080/5090) for local AI cost?
Plan on at least ≈ €2,500 (RTX 5080 16 GB); at least ≈ €4,300 (RTX 5090 24 GB) (at least ≈ $2,300). Prices change quickly — check today's price through the purchase links.
Who is the RTX laptop PC (5080/5090) for local AI designed for?
Run LLMs locally while mobile: nomadic developers, students, consultants.
Go further
Prices in euros (€) are French market prices including VAT, checked by QuelLLM. US prices differ: the Amazon buttons show the current US price.