Glyph 1.1

What Glyph is

Glyph 1.1 is a fine-tuned Llama model built for speed and brevity. It was the first model Sennoric shipped — before Fresco existed — and the lessons from fine-tuning and evaluating it shaped the approach taken with Fresco 1.2.5.

Where it fits

Fresco is the default model — built for general chat, writing, and code. Glyph is smaller and faster, suited to short exchanges where response time matters more than depth.

• Jul 25, 2026Marked deprecated and briefly taken offline while its compute was reallocated to the Project Crucible 600M rerun.
• Jul 30, 2026Back online and answering requests as normal.
• Aug 2026Moved onto dedicated inference capacity separate from Fresco.
• Aug 6, 2026Took the name Glyph. No longer deprecated, and no retirement date set.
• Sep 28, 2026Decommissioned after repeatedly running tool calls despite being explicitly told not to. See the announcement.

Using it

Glyph is decommissioned (see the timeline above) — model: "glyph" now returns an error. Use fresco-1.4, the current default model, instead.

Fresco 1.3 details