ReceiptBench - Structure: leaderboard

The structural parsing field (the list of line items with amounts and tax status) of ReceiptBench (2026), structured extraction of 19 fields from photos and scans of real receipts and invoices (purchase, dining, travel, hotel, taxi and more), 2,000 held-out test images, JSON output. Field-level F1 times 100 (0-100), exact, numeric or list matching with a small LLM judge for near misses; higher is better.

Metric: Structural parsing field (the structured list of line items with amounts and tax status), field-level F1 (times 100, so 0-100) on the 2,000-image held-out test set of ReceiptBench (real travel and expense receipts and invoices, 19 annotated fields), one-shot schema prompt with JSON output; fields are scored by exact, numeric or structured-list matching with a Qwen3-4B semantic judge when string similarity falls below the match threshold; higher is better. Source: arxiv.org. Saturation forecast: Around April 2028. 14 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3 Pro57.81#145
2Qwen 3 VL Plus54.84#512
3InternVL3.5-241B51.12#629
4Qwen 3 VL 4B49.09#1151
5GPT-548.93#178
6Qwen 3 VL 32B48.52#622
7olmOCR-7B48.49
8Qwen 3 VL 8B47.92#917
9PaddleOCR-VL47.46
10InternVL3-78B46.3#759
11InternVL3-8B46.09#1782
12DeepSeek-OCR-small36.37
13InternVL3-2B30.77#3290
14DianJin-OCR-R129.82

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=receiptbench-structure · How It Works · Data refreshed daily, snapshot 2026-11-21.