API Research

Context Window Output Limit Accuarcy Fluency
GPT-4o 124,000 4096
Claude-Haiku 200,000 4096 Moderate Good
Claude-3 Sonnet 200,000 4096 Highest Excellent
Claude-3 Opus 200,000 4096 High Very Good

api 결과 비교

  1. GPT-4o

    조건 1. "You are the assistant who summarizes the given review into 3 sentences and provides 5 keywords.”

    조건 2. "Summarize the text in 3 sentences. Each sentence should be 10 words or fewer. Also, extract 5 key keywords from the content.”

  2. Claude-3 Sonnet

    조건 1. "You are the assistant who summarizes the given review into 3 sentences and provides 5 keywords.”

    조건 2. "Summarize the text in 3 sentences. Each sentence should be 10 words or fewer. Also, extract 5 key keywords from the content.”

  3. Claude-3 Opus

    조건 1. "You are the assistant who summarizes the given review into 3 sentences and provides 5 keywords.”

    조건 2. "Summarize the text in 3 sentences. Each sentence should be 10 words or fewer. Also, extract 5 key keywords from the content.”

최종 데이터 실험 결과_Claude

조건 2. "Summarize the text in 3 sentences. Each sentence should be 10 words or fewer. Also, extract 5 key keywords from the content.”

[gmap_id 1]

[gmap_id 2]

앞으로 해야할 일

[User to Item]