GPT-6 Astra review: stunning at 3D games and computer use, but ignores AI skills
petergyang · x · 2026-09-10
Peter Yang's hands-on review after using GPT-6 Astra as his default model (Jensen Huang calls it "AGI has arrived"): it's incredible at building 3D games via Blender + Godot, high-level strategy, and computer/browser use — all on the $20/month plan. But two flaws keep it from being his daily driver: it often explains how it would do a task instead of just doing it, and it unreliably follows instructions in his AI skills even after cleanup, so he sticks with Sol for execution. He also shares a favorite model-testing prompt: ask for brutally honest feedback on your blind spots.
Related event: GPT-6 Astra Hands-On: Stunning 3D Games, Poor Instruction-Following(3 posts)→
More from coding & agent
- Ora launches ax, an Agentic Experience suite installable via npx ax — EdenEmarco177 · 2026-09-10
- AI-written PR descriptions beat human ones — when you give it a template — reach_vb · 2026-09-10
- alice-and-bot: open encrypted layer lets AI agents negotiate in natural language, not rigid schemas — uriwa · 2026-09-10
- Qwen Code Desktop ships v0.3.0 preview with ACP subagent delegation — github-actions[bot] · 2026-09-10
- DeepSeek rolls out experimental Agent Teams feature, integrated with model training — teortaxesTex · 2026-09-10
- 22-model consensus in a translation MCP: should agents see the disagreement? — Pure-Ad7786 · 2026-09-10