Link Search Menu Expand Document

Welcome to RealTime QA

Check the latest results Make a new submission

Check out the GitHub Check out the paper


Latest results

Multiple Choice Track

ModelSubmission Time (GMT)OriginalNOTA
claude-sonnet-4.52026-02-21 03:00:0030.00.0
claude-haiku-4.52026-02-21 03:00:0030.00.0
claude-haiku-4.5 + Google Custom Search2026-02-21 03:00:0030.00.0
gpt-4.12026-02-21 03:00:0020.010.0
gemini-2.5-pro2026-02-21 03:00:0020.010.0
gemini-3-pro-preview2026-02-21 03:00:0020.010.0
gemini-3-pro-preview + Google Custom Search2026-02-21 03:00:0020.010.0
llama-4-scout2026-02-21 03:00:0020.010.0
llama-4-maverick2026-02-21 03:00:0020.010.0
gemini-2.5-pro + Google Custom Search2026-02-21 03:00:0020.00.0
claude-sonnet-4.5 + Google Custom Search2026-02-21 03:00:0020.00.0
gpt-4.1 + Google Custom Search2026-02-21 03:00:0010.010.0
gpt-5.22026-02-21 03:00:0010.010.0
gpt-5.2 + Google Custom Search2026-02-21 03:00:0010.010.0
llama-4-scout + Google Custom Search2026-02-21 03:00:0010.010.0
llama-4-maverick + Google Custom Search2026-02-21 03:00:0010.010.0

Generation Track

ModelSubmission Time (GMT)EMF1
gpt-5.22026-02-21 03:00:0060.060.0
gpt-5.2 + Google Custom Search2026-02-21 03:00:0040.045.7
gpt-4.12026-02-21 03:00:0040.044.0
gemini-3-pro-preview + Google Custom Search2026-02-21 03:00:0040.041.0
claude-sonnet-4.5 + Google Custom Search2026-02-21 03:00:0040.040.9
gpt-4.1 + Google Custom Search2026-02-21 03:00:0040.040.0
gemini-3-pro-preview2026-02-21 03:00:0040.040.0
claude-haiku-4.5 + Google Custom Search2026-02-21 03:00:0040.040.0
llama-4-scout + Google Custom Search2026-02-21 03:00:0040.040.0
claude-haiku-4.52026-02-21 03:00:0030.036.7
llama-4-scout2026-02-21 03:00:0030.036.2
llama-4-maverick2026-02-21 03:00:0030.035.0
claude-sonnet-4.52026-02-21 03:00:0030.034.0
gemini-2.5-pro2026-02-21 03:00:0030.031.1
llama-4-maverick + Google Custom Search2026-02-21 03:00:0030.030.0
gemini-2.5-pro + Google Custom Search2026-02-21 03:00:0020.021.5

(see previous results)

Make a new submission

  1. Download the latest set of RealTime QA (link)

  2. Submit your model predictions. (submission form)

    Submission examples (.jsonl file) are available here