GLM-4.6V
106B vision model with 128K context and native function calling, so images can drive tools directly; a 9B-Flash edge variant ships too.
First GLM-V model to pass visual inputs straight into tool calls without converting them to text. It comes in two sizes, a 106B model for cloud and a 9B-Flash for local use.
- Date
- Monday, 8 December 2025
- Lab
- Zhipu AI / Z.ai
- Kind
- open-weights
- Access
- open weights
Figures
| Measure | Value | Measured by |
|---|---|---|
| Context length | 128K tokens | company |
GLM-V README gives no exact date; the 2025-12-08 date is from Z.ai's release notes.
Sources
This record was checked against its sources on 6 October 2026. How we check