BestLLMfor Your hardware. Your LLM. Your call.
The Local Copilot Kit APIOpen data Find my LLM
Model fiche

GLM-OCR 1.1B

By Zhipu AI · China

Updated 2026-08-28

vision code small
Parameters
1.1B
License
MIT
Context
128k
VRAM (Q4)
0.6 GB
Released
February 2026

Overview

GLM-OCR (1.1B, Zhipu) is an ultra-compact vision model purpose-built for OCR — document, table, and code extraction — with a 131k context and ~0.6GB VRAM at Q4. Released February 2026.

When to pick this model

  • Document, table, or code extraction pipelines needing a lightweight dedicated OCR model
  • Edge or low-VRAM deployments where a full VLM is overkill
  • Batch document processing where a specialist model outperforms a generalist
  • CPU or integrated-GPU inference

VRAM requirements by quantization

VRAM REQUIRED (GB)Q4_K_M0.6 GBQ5_K_M0.8 GBQ8_01.2 GBFP162.2 GB
QuantizationVRAM required
Q4_K_M (recommended)0.6 GB
Q5_K_M0.8 GB
Q8_01.2 GB
FP16 (no quantization)2.2 GB

VRAM figures include model weights plus a typical 8k KV cache and ~600 MB runtime overhead (Ollama / llama.cpp baseline). Add headroom for higher context lengths.

In practice, GLM-OCR 1.1B fits an 8 GB consumer card at Q4_K_M (0.6 GB). Stepping up to Q8_0 nearly doubles the footprint to 1.2 GB, and unquantized FP16 weights take 2.2 GB — plan your GPU around the Q4 or Q5 figure unless you specifically need the higher fidelity.

Without a GPU, GLM-OCR 1.1B needs roughly 1.4 GB of system RAM to run on CPU via llama.cpp or Ollama — workable for background jobs, but far slower than GPU inference. Throughput estimates from our compatibility engine: around 110 tokens/sec on entry-level GPUs, on the order of 170 tokens/sec on a mid-range card, and up to 220 tokens/sec on high-end hardware — assuming the chosen quantization fully fits in VRAM.

What hardware do you need

The table below matches GLM-OCR 1.1B to common GPU memory tiers, using the highest-fidelity quantization that fully fits each card class. Spilling layers to system RAM works but costs most of the speed, so size your card to the quantization you actually want to run.

GPU memoryExample cardsBest fit for GLM-OCR 1.1B
8 GBRTX 5070 Laptop, RTX 5060, RTX 5060 Ti 8GBFP16 (2.2 GB used)
12 GBRTX 5070, RTX 5070 Ti Laptop, RTX 4080 LaptopFP16 (2.2 GB used)
16 GBRTX 5080, RTX 4080 Super, Radeon RX 9070 XTFP16 (2.2 GB used)
24 GBRTX 4090, Radeon RX 7900 XTX, RTX 5090 LaptopFP16 (2.2 GB used)
32 GBRTX 5090FP16 (2.2 GB used)

Which GPU should you buy to run GLM-OCR 1.1B?

To run GLM-OCR 1.1B locally at Q4, you need ~0.6 GB of VRAM. The best value for this is a RTX 5060 (8 GB VRAM).

Check RTX 5060 price on Amazon →Check RTX 5060 price on Newegg →

As an Amazon Associate, BestLLMfor earns from qualifying purchases, at no extra cost to you. It does not influence our independent rankings.

Strengths

  • Extremely compact (~0.6GB VRAM at Q4)
  • Purpose-built for document OCR
  • Native 128k context
  • Runs on CPU and integrated GPUs

Limitations

  • Not a general-purpose chat model — OCR only
  • Gated weights on Hugging Face
  • Capability ceiling limited by the sub-2B parameter count

Typical workloads

In our catalog grid, GLM-OCR 1.1B is filed under Document OCR, Image Text Extraction, Edge Deployment — the use cases where its size/quality trade-off makes the most sense. Its tags translate to concrete workloads: code generation and review (pair it with an editor integration like Continue.dev or Cline); vision-language work — screenshots, charts, scanned documents.

The 128k-token context window is large enough to hold entire codebases' worth of files or long reports in a single prompt, which is what makes local RAG and document analysis practical. The MIT license is permissive, so shipping it inside a commercial product raises no special legal questions.

Architecture & training

Architecture: Multimodal vision transformer · 0.9B parameters · 128k context · OCR specialization

Training: GLM-OCR family from Zhipu AI / THUDM. Compact model dedicated to text extraction from images and documents.

Verdict

A dedicated, ultra-light OCR specialist for document and code extraction — not a general-purpose model.

Quick start

ollama pull glm-ocr

Or use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.

Similar models worth comparing

Frequently asked questions

How much VRAM does GLM-OCR 1.1B need?

At the recommended Q4_K_M quantization, GLM-OCR 1.1B needs about 0.6 GB of VRAM. Q8_0 takes 1.2 GB, and unquantized FP16 weights take 2.2 GB.

Can GLM-OCR 1.1B run without a GPU?

Yes — with roughly 1.4 GB of system RAM it runs CPU-only through llama.cpp or Ollama. Expect a fraction of GPU speed, which is fine for background or batch jobs but slow for interactive chat.

What context window does GLM-OCR 1.1B support?

GLM-OCR 1.1B supports a 128k-token context window (131,072 tokens).

Can I use GLM-OCR 1.1B commercially?

Yes. GLM-OCR 1.1B is released under MIT, a permissive open-source license that allows commercial use, modification and redistribution.

How fast is GLM-OCR 1.1B on consumer hardware?

Our compatibility engine estimates on the order of 170 tokens/sec on a mid-range GPU and up to 220 tokens/sec on high-end cards, assuming the quantization fully fits in VRAM.

Which quantization of GLM-OCR 1.1B should I download first?

Start with Q4_K_M (0.6 GB) — the standard size/quality sweet spot. Step up to Q5_K_M or Q8_0 only if you have VRAM headroom. It already fits an 8 GB card at FP16.

Tools

Is GLM-OCR 1.1B the right pick for you?

Compute self-hosted ROI → Back to catalog