Can Laptop · GeForce RTX 5090 Laptop · 32 GB RAM run Gemma 4 26B-A4B?
24 GB VRAM + 32 GB RAM · 896 GB/s memory bandwidth · Gemma 4 26B-A4B has 25.8B parameters (3.8B active per token).
Short answer
Yes, excellent for chat and comfortable or better for every workload we test.
Best quantization for chat: Q5_K_M (17 GB of weights) on Ollama. Generation: 120–200 tok/s, leaving about 4.6 GB free for your other apps.
How it rates for each workload
Same model and machine, different jobs. Agents and long documents send far bigger prompts than chat, so they need much more speed and memory.
| Workload | Verdict | Quant | Generation | First reply | Context | |
|---|---|---|---|---|---|---|
| Chatwith Open WebUI | Excellent | Q5_K_M | 120–200 tok/s | under 1 s | 8K | Details |
| Coding questionswith Continue | Excellent | Q5_K_M | 120–190 tok/s | ≈1 s | 16K | Details |
| Coding in a repositorywith Aider | Comfortable | Q6_K | 95–160 tok/s | ≈3 s | 32K | Details |
| Agentic codingwith OpenCode | Comfortable | Q6_K | 90–150 tok/s | ≈3 s | 32K | Details |
| Long documents (64K)with Open WebUI | Excellent | Q5_K_M | 85–150 tok/s | ≈10 s | 64K | Details |
Speeds are bandwidth-based estimates; no direct benchmark exists for this pair. “First reply” is the time to read the workload’s opening prompt. How we calculate
Get it running
Copy-paste commands for the Q5_K_M build with a 64K context. Ollama and llama.cpp download the file straight from Hugging Face.
ollama run hf.co/unsloth/gemma-4-26B-A4B-it-GGUF:UD-Q5_K_XL
Downloads 21.22 GB (UD-Q5_K_XL). Ollama's default context is small: set OLLAMA_CONTEXT_LENGTH=65536 before `ollama serve` for long prompts.
Other models for Laptop · GeForce RTX 5090 Laptop · 32 GB RAM
The best-rated models for chat on this machine.
- Qwen3.6 35B-A3B180–290 tok/sExcellent
- LFM2.5 24B-A2B260–430 tok/sExcellent
- gpt-oss-20b150–260 tok/sExcellent
- Qwen3.5 35B-A3B180–290 tok/sExcellent
- GLM-4.7-Flash (30B-A3B)180–290 tok/sExcellent
- Qwen3 30B-A3B Instruct 2507130–210 tok/sExcellent
Cheapest machines for Gemma 4 26B-A4B
Hardware rated comfortable or better for chat, lowest price first.
- Mac mini M4 24 GB$999 · 25–35 tok/sComfortable
- Mac mini M6 24 GB$1,299 · 30–50 tok/sComfortable
- Laptop · Core Ultra 7 258V (Arc 140V) · 32 GB$1,300 · 15–25 tok/sExcellent
- Laptop · Ryzen AI 9 HX 370 (Radeon 890M) · 32 GB$1,300 · 16–25 tok/sComfortable
- Mac mini M4 32 GB$1,399 · 25–35 tok/sExcellent
- MacBook Air M4 24 GB$1,399 · 25–35 tok/sComfortable
Different tool, context size or workload?
These pages use typical settings. The full check models your exact setup.