Back to Leaderboard
Closed

GPT-5-Nano

OpenAI128K context$0.2/1K pages2026-01-20
Overall Rank
#12
of 16 models
Overall Score
50.7
avg across benchmarks
Best Task
Key Information Extraction
84.7
Weakest Task
Table Understanding
53.9

Benchmark Performance

OlmOCR Benchv1.0
15/16
OverallMathTablePresentAbsentOrder
22.82.455.113.043.147.0
OmniDocBenchv1.5
13/16
OverallText Edit↓Formula CDM↑Table TEDS↑TEDS-S↑Read Order↓
63.40.31961.061.269.50.243
IDP Core Benchv1.0
11/16
OverallKIEOCRTableVQA
65.884.769.645.363.5

Capability Profile

Strength Analysis

Auto-generated from benchmark scores

Strengths

  • Key Information Extraction84.7
  • Text Extraction68.1

Weaknesses

  • Table Understanding53.9
  • Formula61.0