About AigenRig
AigenRig (aigenrig.com) publishes independent, repeatable GPU benchmarks for local AI image and video generation. We measure real ComfyUI workflows (Flux, SDXL, Wan) the same way every time, then explain what that means for builders choosing a card, a prebuilt, a laptop, or a cloud GPU hour.
What we do
- Run a fixed AigenRig Bench Protocol (public methodology) so cards are comparable across reviews.
- Report wall-clock time, peak VRAM, power draw, and cost per 100 generations for local and cloud runs.
- Cover local NVIDIA rigs, Apple Silicon where it is practical, and cloud GPU pricing when local hardware is not the right tradeoff.
Who runs this site
AigenRig is operated by Melis. Contact: see the Contact page.
How we make money
AigenRig may earn commissions or credits when you buy through partner links, including Amazon Associates (US) and cloud GPU referral programs (for example Runpod and Vast.ai). That does not change the price you pay. See our Affiliate Disclosure.
Editorial standards
- Benchmark numbers come from runs under the published protocol, or are clearly labeled as estimates.
- Spec sheets and vendor marketing are never presented as our measured results.
- When a workflow, driver, or model version changes the score, we note it and bump the protocol version if needed.