Pricing
| Type | Price (per 1M) |
|---|---|
| Input tokens | $1.20 |
| Output tokens | $4.00 |
Capabilities
VisionMultimodalReasoningFunction CallingTool UseStructured OutputsCode Execution
About GLM-5V-Turbo
First native multimodal variant of GLM-5 with CogViT visual encoder. Specialized for design-to-code tasks—converts mockups, screenshots, Figma exports, and hand-drawn sketches into HTML, CSS, and JavaScript. Trained with reinforcement learning across 30+ task types with INT8 quantization. Achieved 94.8 on Design2Code benchmark (vs Claude Opus 4.6: 77.3). Supports image, video, and text inputs natively.
Get Started
Model Specs
Released2026-04-01
Parameters744B total, 40B active
Context200k
ArchitectureMixture of Experts