Skip to content

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.

C++

Terminal launch list: Qwen3.6 (own engine path) stays out of Flash-Next's rows

Niko1221979da9d2bca4
diff --git a/strata_cli/launch/catalog.py b/strata_cli/launch/catalog.py
index ebfe0fcb8..2402edc9f 100644
--- a/strata_cli/launch/catalog.py
+++ b/strata_cli/launch/catalog.py
@@ -272,7 +272,10 @@ class Catalog:
vram = comp.main_vram_gb
rec = self.recommendation(hw)
out: list[ModelRow] = []
+ own = self.strata.own_families() # Qwen3.6 is listed for the app; the Terminal's launch list is Flash-Next's
for fam in self.strata.model_table(hw):
+ if fam["family"] in own:
+ continue
for size in fam["sizes"]:
f = fit_on(size, hw.get("ram_gb"), vram)
cat, words = verdict(size.get("on_this_pc"))