GPT-6 Sol generates 3D scenes at four reasoning levels vs Astra
cedric_chee · x · 2026-09-23
User cedricchee shares 3D scenes generated by GPT-6 Sol at four different reasoning levels, with Astra as the reference, linking to a comparison. The post illustrates how output differs across reasoning budgets for 3D generation.
Related event: GPT-6 Sol Generates 3D Scenes Across Reasoning Levels(2 posts)→
More from Multimodal
- Google launches Gemini 3.8 Flash TTS and Flash-Lite TTS, its most expressive audio models — patloeber · 2026-09-23
- Dev open-sources a local Mac studio for testing Qwen-Image-2.1 image generation — janishar · 2026-09-23
- Adobe completes acquisition of Emmy-winning AI enhancer Topaz Labs — chrisfirst · 2026-09-23
- Google DeepMind debuts Gemini 3.8 text-to-speech model — Google DeepMind · 2026-09-23
- Google DeepMind launches Gemini 3.8 TTS with 30-second voice cloning and consent checks — Google DeepMind · 2026-09-23
- Recraft launches V4.1 Flash, generating images in 1.3 seconds as fastest image model — aziz4ai · 2026-09-23