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++
setup: the Mac control path says plainly that the Mac engine has no picture reader yet
Niko12219a22b78887e4
diff --git a/setup.py b/setup.pyindex a0f40b301..60663fc3e 100644--- a/setup.py+++ b/setup.py@@ -6276,7 +6276,7 @@ def mac_control_setup(a) -> int:for n in MF.model_note(model, ram):warn(n)if a.vision in ("yes", "gpu", "cpu"):- warn("images are not part of the app's first release: images off for this model")+ warn("the Mac engine has no picture reader yet: images off for this model (text works as usual)")ctx = a.context or MF.recommend_context(ram)if ctx > MF.TESTED_CONTEXT:say(f" context {ctx}: measured on the Mac up to {MF.TESTED_CONTEXT} (needles found at 8K-32K); longer is kept")