Classement > OpenAI > GPT-6 Luna
GPT-6 Luna
Avis récents: Moyennes · 57 % positives · 24 avis récents
Développeur: OpenAI · Date de sortie: 2026-09-22
- Recommandé · @FrFrederikke: Tak til GPT-6 Luna for Grok ville simpelthen ikke "stigmatisere" den gruppe, som er årsagen til, at vi må sætte koranklodser op, og nægtede derfor at ændre teksten.
- Recommandé · @rishabhguptajs: building one cute personal assistant (similar to Friday of tony stark) for the first time, to see if it really comes out together. for the models: - TTS from Sarvam - LLM is GPT 6 Luna from Openrouter (turns out, Luna is really good and humorous while talking if compared to haik…
- Recommandé · @eric_zwu: I've been building my own design reference library. I want to capture what I find interesting about a design: interactions, details, annotations, etc. That's often what's missing when I feed references to agents, and it's closer to how I internally catalog interesting designs. E…
- Déconseillé · @mschinamodel: 便宜模型的坑：GPT-6 Luna 额度松、价格低，使用量可以完全放开，但改 App 代码准确性差一截。它能改，改完你得回来给它擦屁股——等于给自己请了个爹。省下的钱最后都变成自己加班还回去。
- Déconseillé · @follistatindev: While testing a new subsection of my GameFeel benchmark, something I though could never happen happened, out of hundreds of models Gpt-6-luna scored the worst. it scored negative -5.2% on the new subsection yes negative -5.2% benchmark score, its the only model to score below 0%
- Déconseillé · @Noe1017_Monad: 没想到🙀Haiku已经够差了，gpt-6 luna更烂🥹 就这水平，还不如Qwen3.8 27B不开思考。 价格是对标DeepSeek了，性能对标的应该是Gemini 2.0🫣