👌 𝗗𝗲𝗲𝗽𝗦𝗲𝗲𝗸-𝗥𝟭 𝘃𝘀 𝗖𝗵𝗮𝘁𝗚𝗣𝗧 𝗳𝗼𝗿 𝗞𝗼𝘁𝗹𝗶𝗻 𝗽𝗿𝗼𝗴𝗿𝗮𝗺𝗺𝗶𝗻𝗴 🚀
JetBrains (by Vera Kudrevskaia) has published an analysis on its blog comparing different AI models when working with Kotlin. 🔍
The evaluation included tools like DeepSeek-R1, various OpenAI versions, Anthropic, and Google Gemini
#AndroidDev
JetBrains (by Vera Kudrevskaia) has published an analysis on its blog comparing different AI models when working with Kotlin. 🔍
The evaluation included tools like DeepSeek-R1, various OpenAI versions, Anthropic, and Google Gemini
#AndroidDev
Comments
𝗢𝗽𝗲𝗻𝗔𝗜 𝗼𝟭: 91.93% success rate
𝗗𝗲𝗲𝗽𝗦𝗲𝗲𝗸-𝗥𝟭: 88.82%
𝗢𝗽𝗲𝗻𝗔𝗜 𝗼𝟭-𝗽𝗿𝗲𝘃𝗶𝗲𝘄: 88.82%
𝗢𝗽𝗲𝗻𝗔𝗜 𝗼𝟯-𝗺𝗶𝗻𝗶: 86.96%
𝗢𝗽𝗲𝗻𝗔𝗜 𝗼𝟭-𝗺𝗶𝗻𝗶: 86.34%
𝗚𝗼𝗼𝗴𝗹𝗲 𝗚𝗲𝗺𝗶𝗻𝗶 𝟮.𝟬 𝗙𝗹𝗮𝘀𝗵: 83.23%
𝗔𝗻𝘁𝗵𝗿𝗼𝗽𝗶𝗰 𝗖𝗹𝗮𝘂𝗱𝗲 𝟯.𝟱 𝗦𝗼𝗻𝗻𝗲𝘁: 80.12%
𝗢𝗽𝗲𝗻𝗔𝗜 𝗚𝗣𝗧-𝟰𝗼: 80.12%
𝗢𝗽𝗲𝗻𝗔𝗜 𝗚𝗣𝗧-𝟰𝗼 𝗺𝗶𝗻𝗶: 77.02%
#AndroidDev
47 real questions from the community were evaluated using a rating system (1 to 10), which aligned well with human assessments.
𝗗𝗲𝗲𝗽𝗦𝗲𝗲𝗸-𝗥𝟭: 8.79
𝗢𝗽𝗲𝗻𝗔𝗜 𝗼𝟯-𝗺𝗶𝗻𝗶: 8.70
𝗢𝗽𝗲𝗻𝗔𝗜 𝗼𝟭: 8.62
𝗢𝗽𝗲𝗻𝗔𝗜 𝗼𝟭-𝗽𝗿𝗲𝘃𝗶𝗲𝘄: 8.60
𝗢𝗽𝗲𝗻𝗔𝗜 𝗼𝟭-𝗺𝗶𝗻𝗶: 8.40
𝗢𝗽𝗲𝗻𝗔𝗜 𝗚𝗣𝗧-𝟰𝗼 𝟭𝟭.𝟮𝟬.𝟮𝟬𝟮𝟰: 8.40
𝗔𝗻𝘁𝗵𝗿𝗼𝗽𝗶𝗰 𝗖𝗹𝗮𝘂𝗱𝗲 𝟯.𝟱 𝗦𝗼𝗻𝗻𝗲𝘁: 8.38
𝗢𝗽𝗲𝗻𝗔𝗜 𝗚𝗣𝗧-𝟰𝗼 𝟬𝟴.𝟬𝟲.𝟮𝟬𝟮𝟰: 8.18
𝗔𝗻𝘁𝗵𝗿𝗼𝗽𝗶𝗰 𝗖𝗹𝗮𝘂𝗱𝗲 𝟯.𝟱 𝗛𝗮𝗶𝗸𝘂: 8.01
𝗚𝗼𝗼𝗴𝗹𝗲 𝗚𝗲𝗺𝗶𝗻𝗶 𝟮.𝟬 𝗙𝗹𝗮𝘀𝗵: 7.74
𝗚𝗼𝗼𝗴𝗹𝗲 𝗚𝗲𝗺𝗶𝗻𝗶 𝟭.𝟱 𝗣𝗿𝗼: 7.45
𝗢𝗽𝗲𝗻𝗔𝗜 𝗚𝗣𝗧-𝟰𝗼 𝗺𝗶𝗻𝗶: 7.26
𝗚𝗼𝗼𝗴𝗹𝗲 𝗚𝗲𝗺𝗶𝗻𝗶 𝟭.𝟱 𝗙𝗹𝗮𝘀𝗵: 6.51
𝗚𝗼𝗼𝗴𝗹𝗲 𝗚𝗲𝗺𝗶𝗻𝗶 𝗨𝗹𝘁𝗿𝗮: 6.50
𝗔𝗻𝘁𝗵𝗿𝗼𝗽𝗶𝗰 𝗖𝗹𝗮𝘂𝗱𝗲 𝟯 𝗛𝗮𝗶𝗸𝘂: 6.50
#AndroidDev
https://blog.jetbrains.com/kotlin/2025/02/openai-vs-deepseek-which-ai-understands-kotlin-better/
#AndroidDev #Kotlin #KotlinMultiplatform