Can Desktop PC · Radeon RX 9070 XT · 32 GB RAM run Gemma 4 26B-A4B?
16 GB VRAM + 32 GB RAM · 640 GB/s memory bandwidth · Gemma 4 26B-A4B has 25.8B parameters (3.8B active per token).
Short answer
Yes, excellent for chat; borderline for 64K-token documents.
Best quantization for chat: Q3_K_M (12 GB of weights) on Lemonade. Generation: 170–240 tok/s, leaving about 2.0 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 | Q3_K_M | 170–240 tok/s | under 1 s | 8K | Details |
| Coding questionswith Continue | Comfortable | Q3_K_M | 170–230 tok/s | ≈2 s | 16K | Details |
| Coding in a repositorywith Aider | Acceptable | Q3_K_M | 150–210 tok/s | ≈4 s | 32K | Details |
| Agentic codingwith OpenCode | Acceptable | Q3_K_M | 150–200 tok/s | ≈5 s | 32K | Details |
| Long documents (64K)with Open WebUI | Borderline | Q3_K_M | 110–150 tok/s | ≈15 s | 64K | Details |
Speeds are estimates calibrated against measured benchmarks on similar hardware. “First reply” is the time to read the workload’s opening prompt. How we calculate
Get it running
Copy-paste commands for the Q3_K_M build with a 32K context. Ollama and llama.cpp download the file straight from Hugging Face.
ollama run hf.co/unsloth/gemma-4-26B-A4B-it-GGUF:UD-Q3_K_XL
Downloads 12.91 GB (UD-Q3_K_XL). Ollama's default context is small: set OLLAMA_CONTEXT_LENGTH=32768 before `ollama serve` for long prompts.
Other models for Desktop PC · Radeon RX 9070 XT · 32 GB RAM
The best-rated models for chat on this machine.
- gpt-oss-20b160–210 tok/sExcellent
- MiniCPM5 2B160–220 tok/sExcellent
- Qwen3.5 9B55–75 tok/sExcellent
- Qwen3.5 4B100–130 tok/sExcellent
- Gemma 4 12B45–60 tok/sExcellent
- GLM-4.6V-Flash (10B)50–70 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.