Running a 125B Qwen3.8 model on a GPU with 12GB VRAM: Strata debuts, achieving 94 tokens/second on a single RTX 5070
Tech media gigazine reported today (October 6) that developer Niko1221 open-sourced the Strata engine, which can run the quantized Qwen3.8-Flash-Next model (125 billion parameters) on consumer graphics cards with 12GB or more VRAM.
Return to original text
A short introduction is retained here; the full text from the source website is not reposted. For specific wording and subsequent updates, please refer to the original article.
Open the original article on ITHome