HP · Ryzen AI Max+ PRO 395 · Laptop
Best AI Models for HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB)
Runs 14B and 120B MoE models well; 27–31B and 70B fit but are slow (10 and 4 tok/s).
| # | Model | Speed | llmrun Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Qwen3.6 35B A3BAlibabaHow to run → Q4_K_M · 22.0 GB · 46.7 t/s | 46.7 tok/s | Q4_K_M · 22.0 GB | |
| 2 | Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 51.3 t/s | 51.3 tok/s | Q4_K_M · 16.5 GB | |
| 3 | Qwen3 30B A3B Thinking 2507Alibaba Q4_K_M · 19.4 GB · 58.3 t/s | 58.3 tok/s | Q4_K_M · 19.4 GB | |
| 4 | GLM 4.5 AirZ.ai Q4_K_M · 67.6 GB · 17.0 t/s | 17.0 tok/s | Q4_K_M · 67.6 GB | |
| 5 | Q4_K_M · 19.3 GB · 72.0 t/s | 72.0 tok/s | Q4_K_M · 19.3 GB |
| # | Model | Speed | Coding Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Qwen3.6 35B A3BAlibabaHow to run → Q4_K_M · 22.0 GB · 46.7 t/s | 46.7 tok/s | Q4_K_M · 22.0 GB | |
| 2 | Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 51.3 t/s | 51.3 tok/s | Q4_K_M · 16.5 GB | |
| 3 | GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 55.7 t/s | 55.7 tok/s | Q4_K_M · 13.1 GB | |
| 4 | GPT OSS 120BOpenAI Q4_K_M · 70.8 GB · 40.9 t/s | 40.9 tok/s | Q4_K_M · 70.8 GB | |
| 5 | Q4_K_M · 19.3 GB · 72.0 t/s | 72.0 tok/s | Q4_K_M · 19.3 GB |
| # | Model | Speed | Reasoning Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Qwen3.6 35B A3BAlibabaHow to run → Q4_K_M · 22.0 GB · 46.7 t/s | 46.7 tok/s | Q4_K_M · 22.0 GB | |
| 2 | Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 51.3 t/s | 51.3 tok/s | Q4_K_M · 16.5 GB | |
| 3 | Qwen3.5 9BAlibaba Q4_K_M · 6.6 GB · 26.2 t/s | 26.2 tok/s | Q4_K_M · 6.6 GB | |
| 4 | GPT OSS 120BOpenAI Q4_K_M · 70.8 GB · 40.9 t/s | 40.9 tok/s | Q4_K_M · 70.8 GB | |
| 5 | Q4_K_M · 19.3 GB · 72.0 t/s | 72.0 tok/s | Q4_K_M · 19.3 GB |
| # | Model | Speed ↓ | llmrun Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Q4_K_M · 19.4 GB · 58.3 t/s | 58.3 tok/s | Q4_K_M · 19.4 GB | |
| 2 | Q4_K_M · 19.3 GB · 72.0 t/s | 72.0 tok/s | Q4_K_M · 19.3 GB | |
| 3 | Qwen3 30B A3B Instruct 2507Alibaba Q4_K_M · 19.4 GB · 58.3 t/s | 58.3 tok/s | Q4_K_M · 19.4 GB | |
| 4 | GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 55.7 t/s | 55.7 tok/s | Q4_K_M · 13.1 GB | |
| 5 | Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 51.3 t/s | 51.3 tok/s | Q4_K_M · 16.5 GB |
| # | Model | Speed | Agents Score | Quant · Memory |
|---|---|---|---|---|
| 1 | GPT OSS 120BOpenAIHow to run → Q4_K_M · 70.8 GB · 40.9 t/s | 40.9 tok/s | Q4_K_M · 70.8 GB | |
| 2 | Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB · 46.7 t/s | 46.7 tok/s | overall | Q4_K_M · 22.0 GB |
| 3 | Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 51.3 t/s | 51.3 tok/s | overall | Q4_K_M · 16.5 GB |
| 4 | Q4_K_M · 19.3 GB · 72.0 t/s | 72.0 tok/s | Q4_K_M · 19.3 GB | |
| 5 | GLM 4.7 FlashZ.ai Q4_K_M · 19.9 GB · 44.9 t/s | 44.9 tok/s | Q4_K_M · 19.9 GB |
The HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB) is a 128 GB laptop with more memory than speed: it runs 14B and 120B MoE models well. The system keeps about 32 GB, leaving 96 GB for models. Prompt reading is its weak spot: a 4K-token prompt waits 4.4 s before the first word on Qwen3.6 35B A3B. At its $3,500 MSRP it gives 13.3 tok/s per $1,000, 6th of the 8 machines compared here. Ryzen AI Max+ PRO 395, released 2025.
HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB) strengths and weaknesses
Strengths
- Runs 120B MoE models: GPT OSS 120B at 40.9 tok/s in 70.8 GB
- Low power draw: 55 W at full load for the whole machine
Weaknesses
- Slow on large dense models: Llama 3.3 70B Instruct at 3.6 tok/s, Qwen3.8 27B at 9.6 tok/s
- ROCm, not CUDA: No FP8 weights; MXFP4 only through llama.cpp, not vLLM
- Same chip for less: HP Z2 Mini G1a (Ryzen AI Max+ PRO 395, 128 GB) uses the same Ryzen AI Max+ PRO 395 at $3,342 MSRP
Memory for models
96 GB of 128 GBEverything on the size ladder fits. Qwen3.6 35B A3B and Gemma 4 26B A4B IT together use 38.5 GB and leave 57.5 GB free.
Speed per dollar and per watt
On Qwen3.6 35B A3B (35B params, 3B active) at Q4_K_M. New means a shop price, either a street quote or the maker's own price; used is compared with used. The last two columns are the speed you get per $1,000 of price and per 100 W of power; higher is better. The HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB) has no current new price, so it is shown at its MSRP.
| Machine | New price | tok/s | tok/s per $1,000 | tok/s per 100 W |
|---|---|---|---|---|
Mac Studio M3 Ultra96 GB | about $3,999 | 76.6 | 19.2 | 28.4 |
Mac Studio M4 Max128 GB | about $3,499 | 64.2 | 18.3 | 44.3 |
| about $3,342 | 46.7 | 14.0 | 38.9 | |
Mac Studio M2 Ultra192 GB | about $5,599 | 75.9 | 13.6 | 25.7 |
DGX Spark128 GB | about $3,999 | 54.0 | 13.5 | 22.5 |
| about $3,500 | 46.7 | 13.3 | 84.9 | |
MacBook Pro 16" M5 Max128 GB | about $5,399 | 67.8 | 12.6 | – |
Mac Studio M3 Ultra256 GB | about $7,099 | 76.6 | 10.8 | 28.4 |
| Machine | Used price | tok/s | tok/s per $1,000 | tok/s per 100 W |
|---|---|---|---|---|
RTX 409024 GBBest used value | about $3,000 | 151 | 50.3 | 33.6 |
RTX 509032 GB | about $6,400 | 213 | 33.3 | 37.1 |
Mac Studio M3 Ultra96 GB | about $4,525 | 76.6 | 16.9 | 28.4 |
Cost per token: the HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB) against 24–256 GB machines and GPUs
On Qwen3.6 35B A3B (35B params, 3B active, Q4_K_M, 4K context). Speed, cost per token and power per token depend on the model.
- Machine, MSRP
- $4.75
- Power
- $0.05
Renting costs as much as buying after ≈3,500 hours (≈2.4 years at 4 h a day).
Each dot is a machine or GPU; hover or tap it for its numbers, click to open it. Machines without a price for the chosen basis are left out. Machines without a published power figure are left out of Per watt and costed without power. How we calculate cost per token
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.
Output speed depends on memory bandwidth (256 GB/s). Grey rows need more than the 96 GB available.
- 9BQwen3.5 9B · Q4_K_M · 6.6 GB26.2tok/s
- 14BQwen3 14B · Q4_K_M · 11.9 GB17.5tok/s
- 9.6tok/s
- 35B MoEQwen3.6 35B A3B · Q4_K_M · 22.0 GB46.7tok/s
- 3.6tok/s
- 120B MoEGPT OSS 120B · Q4_K_M · 70.8 GB40.9tok/s
Estimated speeds: Q4_K_M, 4K context, one user. How we estimate
Unified-memory platforms: writing vs compute
- Orin64 GB · 204.8 GB/s41.743
- Ryzen AI Max+ 395128 GB · 256 GB/s46.730
- GB10128 GB · 273 GB/s54.0125
- Jetson T5000 (Thor)128 GB · 273 GB/s54.0259
- M4 Pro64 GB · 273 GB/s43.217
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 1270 models on the HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB)
Featured onlyBest quant that fits in 96 GB, ordered by llmrun Score. Fit grades how well it runs here, S best. How ranking works
| Model | Quant | VRAM | Speed | Context | Score | Fit |
|---|---|---|---|---|---|---|
NVIDIA Nemotron 3.5 Super VL 120B A12B BF16VisionFunctions | Q4_K_M | 87.7 GB | 19.1 t/s | — | C43, Usable | |
JEV 27B VLVisionFunctions | Q4_K_M | 17.9 GB | 9.6 t/s | 262K | B47, Decent | |
Ornith 1.5 122BChatFunctions | Q4_K_M | 75.5 GB | 19.6 t/s | 262K | B60, Decent | |
Ornith 1.5 35B A3BChatFunctions | Q4_K_M | 22.0 GB | 46.7 t/s | 262K | A78, Runs well | |
MiniCPM5 2BChatFunctions | Q4_K_M | 2.5 GB | 87.6 t/s | 131K | S89, Runs great | |
MiMo V2.6 Distill Qwen 9BVisionCodeFunctions | Q4_K_M | 6.5 GB | 26.8 t/s | 262K | A68, Runs well | |
Qwen3.5 4BVisionFunctions | Q4_K_M | 3.4 GB | 51.0 t/s | 262K | A80, Runs well | |
Laguna S 2.1ChatFunctions | Q4_K_M | 72.5 GB | 28.2 t/s | 1049K | A69, Runs well | |
Hy Memory 30B A3BChatFunctions | Q4_K_M | 19.1 GB | 63.1 t/s | 262K | A84, Runs well | |
Laguna XS 2.1ChatFunctions | Q4_K_M | 21.3 GB | 66.2 t/s | 262K | S85, Runs great | |
Laguna XS.2ChatFunctions | Q4_K_M | 21.3 GB | 66.2 t/s | 262K | S85, Runs great | |
Spark X2.6 4BChatFunctions | Q4_K_M | 4.3 GB | 56.2 t/s | 1049K | A82, Runs well | |
Mistral Medium 3.5 128BChatFunctions | Q4_K_M | 82.8 GB | 2.1 t/s | 262K | C30, Usable | |
GLM 4.6VVisionFunctions | Q4_K_M | 66.0 GB | 17.7 t/s | 131K | B58, Decent | |
Qwen3.5 9BVisionFunctions | Q4_K_M | 6.6 GB | 26.2 t/s | 262K | A67, Runs well |
Specs
- Chip
- Ryzen AI Max+ PRO 395 · Radeon 8060S
- CPU cores
- 16
- GPU cores
- 40
- Unified memory
- 128 GB LPDDR5x-8000 (256-bit)
- For models
- 96 GB (The system keeps 32 GB)
- Memory bandwidth
- 256 GB/s
- FP16 (dense)
- 29.7 TFLOPS
- NPU
- XDNA 2
- Power
- 55 W max
- Size
- 14-inch mobile workstation
- Released
- 2025-03-01
- MSRP
- $3,500
HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB) FAQ
Is the HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB) worth it for local LLMs in 2026?
At its $3,500 MSRP it gives 13.3 tok/s per $1,000 on Qwen3.6 35B A3B, 6th of the 8 machines compared here. It runs 14B and 120B MoE models well; 27–31B and 70B fit but are slow (10 and 4 tok/s).
How much memory is available for AI on the HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB)?
About 96 GB of the 128 GB. The system keeps roughly 32 GB for itself and other apps.
What is the largest model the HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB) can run?
Mistral Medium 3.5 128B (127.7B) is the largest at 4-bit or better: at Q4_K_M it uses 82.8 GB, 86% of the available memory, and runs at about 2.1 tok/s.
Can the HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB) run Qwen3.6 35B A3B?
Yes. Qwen3.6 35B A3B fits at Q4_K_M, using 22.0 GB of the 96 GB available, and it is the best model that runs well on this machine.
Can the HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB) run 27–31B models?
Yes, but slowly: Qwen3.8 27B at Q4_K_M uses 17.9 GB and writes about 9.6 tok/s.
How long does the HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB) take to read a long prompt?
On Qwen3.6 35B A3B, about 4.4 s for a 4K-token prompt and 35 s for a 32K-token document before the first word appears.
Is renting a GPU cheaper than buying the HP ZBook Ultra G1a 14 (Ryzen AI Max+ PRO 395, 128 GB)?
An RTX PRO 6000 Blackwell Max-Q Workstation Edition (96 GB) rents from $1.00/hr. Renting costs as much as buying after ≈3,500 hours (≈2.4 years at 4 h a day).