AI-Modell-Intelligence-Plattform
← Benchmarks

Beste KI-Modelle im Schlussfolgerung-Benchmark

Ranking von KI-Modellen nach Schlussfolgerung-Benchmark-Score.

RangModellAnbieterPunktzahlDatensatzVersionDatum
1DeepSeek R1DeepSeek71.5gpqapaper-20252025-01-20
2Claude 3.5 SonnetAnthropic59.4gpqaofficial-20242024-06-20
3GPT-4oOpenAI53.6gpqaofficial-20242024-05-13
4Llama 3.1 405BMeta51.1gpqaofficial-20242024-07-23
5Claude 3 OpusAnthropic50.4gpqaofficial-20242024-03-04
6Gemini 1.5 ProGoogle46.5gpqatechreport-20242024-05-14
7GPT-4o miniOpenAI40.2gpqaofficial-20242024-07-18
8Claude 3 HaikuAnthropic33.3gpqaofficial-20242024-03-04

Source

Datensatz
gpqa
Version
paper-2025
Datum
2025-01-20

Benchmark

Related Resources

LeaderboardRelated RankingBenchmarksBenchmarks
Featured Models
DeepSeek R1 (71.5) →Claude 3.5 Sonnet (59.4) →GPT-4o (53.6) →
Use Cases
Reasoning →