Rune — local AI model by Surogate
Written by Jakub Rusinowski · Last updated
Surogate's decision model on Gemma 4 26B-A4B (mixture-of-experts, text and images). The highest-scoring downloadable model on the community Decision Index 0.2.1 snapshot of 28 September 2026.
Variants
The smallest Rune variant needs about 16 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache.
| Model | VRAM at Q4 | VRAM | Context | Run it |
|---|---|---|---|---|
| Rune 26B-A4B v3 → 25.8B (4B active) | ~16.4 GB | Not published | surogate serve |
Memory is quantized weights plus overhead at Q4_K_M, from the same engine as the GPU & VRAM checker.
How to run Rune locally
Install Ollama, then pull the tag.
Served by surogate serve. This model does not run in Ollama.
Pick a size above for its own VRAM figure, speed estimate and install command.
Licence
Commercial use permitted. No usage restrictions beyond attribution.
Applies to: Rune 26B-A4B v3Recommended GPU
The cheapest catalogued GPU that runs Rune locally (min 16 GB VRAM) is the AMD Radeon RX 7900 XT (20 GB).
Rune — frequently asked questions
How much VRAM does Rune need?
Rune needs about 16 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: Rune 26B-A4B v3 (16 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.
Can I run Rune on an RTX 4090 (24 GB)?
Yes — Rune runs on an RTX 4090 (24 GB) and other 24 GB cards such as the RTX 3090. Smaller variants also fit comfortably on 8–16 GB GPUs at Q4_K_M.
What quantization should I use for Rune?
Q4_K_M is the best balance of quality and VRAM for Rune in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.
How do I run Rune locally?
Rune is a decision model: it is called over an HTTP endpoint (/v1/systemone), not chatted with. Where a variant is available in Ollama 0.35 or newer, pull it with `ollama pull` and send requests to the local server; the others ship their own server. Each variant page shows the exact commands for that model.