ReceiptBench - Perception: leaderboard

Basic perception fields (raw times, totals, currency clues, invoice and tax numbers, seller name) 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: Basic perception fields (raw times, totals, currency clues, invoice and tax numbers, seller name), 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 February 2027. 14 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3 Pro73.6#145
2Qwen 3 VL Plus73.06#512
3GPT-573.04#178
4InternVL3.5-241B68.53#629
5Qwen 3 VL 8B66.64#917
6Qwen 3 VL 32B66.45#622
7InternVL3-78B64.86#759
8Qwen 3 VL 4B64.07#1151
9PaddleOCR-VL61.33
10olmOCR-7B57.36
11InternVL3-8B56.96#1782
12DianJin-OCR-R150.38
13InternVL3-2B38.51#3290
14DeepSeek-OCR-small36.88

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

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