13,353 handwritten text lines from 657 writers. Standard handwriting benchmark.
Wer is the reported evaluation metric for IAM. Codesota tracks published model scores on this metric so readers can compare state-of-the-art results across sources and model families.
Lower is better
Muted rows were not state of the art when published — an earlier or same-year result already scored better.
| Rank | Model | Trust | Score | Year | Links | Fix |
|---|---|---|---|---|---|---|
| 01 | HTR-ConvText | verified | 12.9 | 2026 | Source ↗ | Looks wrong? |
| 02 | HTR-VT | verified | 14.9 | 2026 | Source ↗ | Looks wrong? |
| 03 | VAN | verified | 16.3 | 2026 | Source ↗ | Looks wrong? |
Cer is the reported evaluation metric for IAM. Codesota tracks published model scores on this metric so readers can compare state-of-the-art results across sources and model families.
Lower is better
Muted rows were not state of the art when published — an earlier or same-year result already scored better.
| Rank | Model | Trust | Score | Year | Links | Fix |
|---|---|---|---|---|---|---|
| 01 | TrOCR-large | verified | 2.89 | 2026 | Source ↗ | Looks wrong? |
| 02 | TrOCR-base | verified | 3.42 | 2026 | Source ↗ | Looks wrong? |
| 03 | HTR-ConvText | verified | 4.00 | 2026 | Source ↗ | Looks wrong? |
| 04 | HTR-VT | verified | 4.70 | 2026 | Source ↗ | Looks wrong? |
| 05 | VAN | verified | 5.00 | 2026 | Source ↗ | Looks wrong? |