Every week we run new open-weight models on real tasks (extraction, coding, summarization, agentic tool use) on consumer GPUs we own, and publish measured speed, quality and failure examples. No vendor-supplied numbers.
| Model | Quant | GPU | Task | tok/s | Score |
|---|---|---|---|---|---|
| example-model-A | Q4_K_M | example GPU | Invoice extraction | — | — |
| example-model-B | Q8_0 | example GPU | Python bugfix | — | — |