TrOCR
Transformer-based OCR that reads handwritten and printed text lines.
About
TrOCR is Microsoft's Transformer encoder-decoder model for optical character recognition, introduced in 2021. It pairs a vision Transformer encoder with an autoregressive text decoder to transcribe single lines of handwritten or printed text.
It ships in small, base, and large sizes across printed and handwritten domains (62M to 558M parameters), all on Hugging Face as microsoft/trocr-*. The base checkpoints are a common default for handwriting pipelines and run locally with no API needed. Release timing below is era-level; consult the paper for the exact date.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 110; Trending score 78/100 from researcher assessment.