table-transformer-structure-recognition vs table-transformer-structure-recognition-v1.1-all

Specs, VRAM requirements and download trends — updated 18 October 2026.

microsoft · A · Open weights
Params
30 M
Context
1,024
30d
1.3 M
mit · commercial OK
microsoft · B · Open weights
Params
30 M
Context
—
30d
200 K
mit · commercial OK

Specification comparison

Differences are highlighted; identical values are muted.

Specification table-transformer-structure-recognition table-transformer-structure-recognition-v1.1-all
Parameters 30 M 30 M
Architecture TableTransformerForObjectDetection TableTransformerForObjectDetection
Context length 1,024 —
Licence mit mit
Commercial use Allowed Allowed
Languages — —
Downloads 30d 1,275,589 199,823
Downloads all time 39.9 M 18.6 M
Quantizations on the Hub SAFETENSORS SAFETENSORS
Gated No No
First seen on the Hub 2022-10-14 2023-11-18

Download trend

Daily snapshots, last 89 days (28 Jul – 24 Oct)

1.6 M200 K
table-transformer-structure-recognition table-transformer-structure-recognition-v1.1-all

VRAM side by side

On RTX 4090 · 24 GB · 8K context unless noted

table-transformer-structure-recognition · fp16 0.6 GB / 24 GB
✅ Runs comfortably
table-transformer-structure-recognition-v1.1-all · fp16 0.6 GB / 24 GB
✅ Runs comfortably

Adjacent comparisons