OpenAI Reportedly Developing 'Mona Lisa' and 'Luna Lisa' Image Models
mark_k · x · 2026-08-23
Leaks suggest OpenAI is developing two new image models: the larger "Mona Lisa-1" (potentially GPT-Image-2.5) offers noticeable but not spectacular improvements over the current version. The smaller "Luna Lisa Alpha" (based on GPT Luna) is faster and performs on par with existing models. Unfortunately, both reportedly still suffer from noise artifact issues.
More from Models
- Anthropic to watermark all Claude output using invisible SynthID — every · 2026-08-23
- Arena's mystery model "adamant-ananke" confirmed to be GLM with post-training magic — ChrisGPT · 2026-08-23
- DeepSeek's Flash Vision Beats Luna at a Quarter of the Price — bindureddy · 2026-08-23
- Qwen3.8 27B Uncensored GGUF Released with 262k Context and Vision Support — BLUECOW009 · 2026-08-23
- Benchmarking Gemini 3.7/3.6 Flash Coding Capabilities — YogurtNo349 · 2026-08-23
- Qwen 3.8 27B hits 91.9 median TPS and 99 fastest TPS — gajesh · 2026-08-23