Python

deprecatedmit900M

GLM-OCR is a lightweight professional OCR model with parameters as small as 0.9B, yet it achieves state-of-the-art performance across multiple capabilities. It sets a new benchmark for document parsing with its “small size and high accuracy.” Key features include:

Context
200K
Max output
128K

Capabilities

Reasoning & Interaction
Reasoning
Tool Calling
Infrastructure
Streaming
Prompt Caching
Fine-tuning

Details

Creator zai
Family—
Licensemit
Statusdeprecated
Type—
Input modalitiestext
Output modalitiestext
Open weightYes
Parameters900M
Architecture—
Context window200K tokens
Max input—
Max output128K tokens
Reasoning tokens—
Knowledge cutoff—
Release date—
Sourceofficial

Tools

Function Callingfunction_calling
Call external functions and APIs

API

GET/v1/models/zai/python

Python

deprecated

GLM-OCR is a lightweight professional OCR model with parameters as small as 0.9B, yet it achieves state-of-the-art performance across multiple capabilities. It sets a new benchmark for document parsing with its “small size and high accuracy.” Key features include:

Changes · 6 entries
pythonupdateca7a400Nov 2, 2026, 03:02 AM
page_urlhttps://docs.z.ai/guides/image/cogview-4→https://docs.z.ai/guides/vlm/glm-ocr
modalitiesoutput["image"]→["text"]
capabilitiestool callyesreasoningyesprompt cachingyes
descriptionGLM-OCR is a lightweight professional OCR model with para...
taglineCogView-4 is Z.AI’s first open-source text-to-image model.
context_window200000
max_output_tokens128000
tools["function_calling"]
parameters0.9
statusdeprecated
pythonupdateb2f45bfOct 29, 2026, 03:29 AM
statusdeprecated
pythonupdatee3d7dedAug 31, 2026, 04:13 AM
page_urlhttps://docs.z.ai/guides/vlm/glm-5v-turbo→https://docs.z.ai/guides/vlm/glm-ocr
descriptionGLM-5V-Turbo is Z.AI's first multimodal coding foundation...→GLM-OCR is a lightweight professional OCR model with para...
parameters0.9
pythonupdate3cddff5Aug 13, 2026, 02:54 AM
page_urlhttps://docs.z.ai/guides/image/cogview-4→https://docs.z.ai/guides/vlm/glm-5v-turbo
descriptionCogView-4 is Z.AI’s first open-source text-to-image model...→GLM-5V-Turbo is Z.AI's first multimodal coding foundation...
context_window200000
max_output_tokens128000
tools["function_calling"]
pythonupdateab879c8Jun 18, 2026, 08:32 AM
descriptionCogView-4 is Z.AI’s first open-source text-to-image model...
taglineCogView-4 is Z.AI’s first open-source text-to-image model.
pythoncreate2b100acMar 23, 2026, 05:30 AM