glm-ocr

Common Name: GLM-OCR

ChatGLM
Released on Feb 3 12:00 AM
Compare

Zhipu AI's 0.9B document OCR model (94.62 on OmniDocBench V1.5). Converts images and PDFs to Markdown while preserving table, formula, and layout structure. Served through the layout parsing endpoint, not chat completions; a single call accepts one image (≤10MB) or PDF (≤50MB, ≤100 pages).

Specifications

Context
131.1K
Inputimage, pdf
Outputtext

Performance (7-day Average)

Collecting…
Collecting…
Collecting…

Pricing

Input¥0.22/MTokens
Output¥0.22/MTokens

Availability Trend (24h)

Performance Metrics (24h)

Similar Models

¥0.11/¥0.11/M
ctx128Kmax16Kavailtps
InOutCap

Zhipu AI's fastest GLM-4 variant optimized for high-throughput inference.

¥1.10/¥3.30/M
ctx128Kmax32Kavailtps
InOutCap

Zhipu AI's vision-reasoning model. Processes text, images, video, and files with strong front-end code replication and GUI analysis capabilities.

¥0.165/¥1.65/M
ctx128Kmax32Kavailtps
InOutCap

Lightweight, high-speed variant of GLM-4.6V with multimodal tool calling and long-context visual reasoning.

Free/Free
ctx128Kmax32Kavailtps
InOutCap

Free variant of GLM-4.6V for cost-sensitive multimodal applications.