GLM-OCR
GLM-OCRCompare inonOmniDocBench full v1.6
Alternatives?
ModelScoreCostPer taskWhat changes
CheaperPaddleOCR VL-1%-90%$0.02 / $0.02XOpenTeleOCR+2%——
CheaperDeepSeek OCR 2Open-5%-85%$0.03 / $0.03AOpenOvisOCR2+1%+162%$0.30 / $1.20POpenPaddleOCR VL 1.6+1%——ModelScoreCostPer taskWhat changes
PaddleOCR VL-1%-90%$0.02 / $0.02DeepSeek OCR 2Open-5%-85%$0.03 / $0.03ModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
XTeleOCROpen+2%——AOvisOCR2Open+1%+162%$0.30 / $1.20PPaddleOCR VL 1.6Open+1%——PPaddleOCR VL 1.5Opensame——Unlimited OCROpen-1%——ModelScoreCostPer taskWhat changes
GLM 5.3 FlashXOpenn/a—$0.37 / $1.25Memory (4-bit): not verifiedGLM 5.3 FlashmaxOpenn/a—$0.28Memory (4-bit): 3 GB → 193 GB+5GLM 5.3lowOpenn/a—$0.85Memory (4-bit): 3 GB → 452 GB+5GLM 5.2non-thinkingOpenn/a—$1.40 / $4.40Memory (4-bit): 3 GB → 452 GB+1GLM 5.1thinkingOpenn/a—$1.02Memory (4-bit): 3 GB → 452 GB+1GLM 5V Turbothinkingn/a—$1.20 / $4GLM 4.6Vnon-thinkingOpenn/a—$0.30 / $0.90Memory (4-bit): 3 GB → 65 GBGLM-4.6V-FlashOpenn/a—$0 / $0Memory (4-bit): 3 GB → 6.2 GBAll Z.ai models →Run it yourself
estimateFits on✓ 8 GB (phone, laptop) · compressed✓ 16 GB · compressed✓ 24 GB GPU · compressed✓ 48 GB GPU · compressed✓ 80 GB GPU · compressed
LicenceMITfree for commercial use
Benchmarks
in Vision & DocumentsOmniDocBench · v1.6 · full95.2#4
FAQ
How much does GLM-OCR cost?›
$0.20 per 1M input tokens and $0.20 per 1M output tokens; cheapest via lab-z-ai at $0.03 / $0.03.
Is GLM-OCR free to use?›
There is no free API; the cheapest paid price is $0.03 per 1M input tokens. The weights are open, so you can run it yourself.
How good is GLM-OCR, and what is it best at?›
#4 of 22 in Vision & Documents.
Is GLM-OCR better than PaddleOCR VL 1.6?›
On OmniDocBench · v1.6 · full, GLM-OCR scores 95.2 vs 96.3 — lower.
How can I use GLM-OCR?›
Through the Z.ai API.
Is GLM-OCR open source?›
Yes, the weights are open (MIT licence).
Can I run GLM-OCR locally?›
Yes, with about 3 GB of GPU or unified memory.
Can GLM-OCR read images?›
It accepts text and images, PDF files as input.
What is GLM-OCR?›
GLM-OCR is an open-weights language model from Z.ai.
What are the best alternatives to GLM-OCR?›
PaddleOCR VL (cheaper), TeleOCR (open weights).