GLM-Image

It is a text-to-image and image-to-image generation model employing a hybrid architecture combining an autoregressive generator and a diffusion decoder. It excels in generating high-fidelity images with precise text rendering and semantic understanding, particularly in complex, information-dense scenarios.

Key featuers:

• The model supports diverse tasks, including text-to-image generation, image editing, style transfer, and multi-subject consistency preservation.
• Image resolutions must be divisible by 32 (e.g., 32×32, 1024×1024). 
• High GPU memory usage (~23GB) recommended with CPU-offloading (in cost of reduces inference speed).
• Released under the MIT License. Includes components (VQ tokenizer, VIT weights) under Apache-2.0 licensing. 

---

The model is a component of the image generation pipeline, consisting of:

• Text encoder: ~0.2B parameters,
• Vision language encoder: ~10B parameters,
• Transformer: ~7B parameters,
• VAE: ~406M parameters,

Total: ~17.7B parameters

Announce Date: 08.01.2026
Parameters: 7B
Developer: Z.ai
License: MIT

Use our pre-built public endpoints for free to test inference and explore GLM-Image capabilities. You can obtain an API access token on the token management page after registration and verification.

Rent your own physically dedicated instance with hourly or long-term monthly billing.

We recommend deploying private instances in the following scenarios:

• maximize endpoint performance,
• enable full context for long sequences,
• ensure top-tier security for data processing in an isolated, dedicated environment,
• use custom weights, such as fine-tuned models or LoRA adapters.

Contact our dedicated neural networks support team at nn@immers.cloud or send your request to the sales department at sale@immers.cloud.