[CLSTR]
< Back to all clusters
[TECHNOLOGY] · 3 sources

started · updated

Part of situation: Strata engine for local large language model inference (3 clusters)

Strata engine enables large AI models on consumer gaming PCs

Strata inference engine Qwen3. 8-Flash-Next open-source AI large language models local LLM execution

The open-source inference engine Strata, developed by Niko1221, enables the execution of the 125-billion-parameter Qwen3. 8-Flash-Next large language model on consumer-grade gaming hardware. The engine works by spreading the model across a user's GPU, CPU, RAM, and SSD, allowing it to run on machines equipped with at least 12 GB of VRAM and 32 GB of RAM.

In performance testing, the engine achieved speeds of up to 120 tokens per second on hardware such as the Radeon 7900 XTX. The system is designed for complete privacy and local operation, ensuring no data leaves the user's machine and eliminating per-token cloud costs. The latest v0. 1. 38 release includes a security update to block DNS-rebinding and cross-site attacks against its local server by rejecting requests from untrusted hosts.

Entities

GitHub · Niko1221 · Qwen3. 8-Flash-Next · Strata

Sources

GitHub - Niko1221/Strata: Qwen3. 8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input. [github.com]
A Free Installer Turns Gaming PC Into Local AI Servers [www.electronicsforu.com]
Strata v0. 1. 38 Runs a 125-Billion-Model LLM on Any Gaming PC [www.linuxcompatible.org]