User Tests OpenAI's New Model and Platform
Normal-End1169 · reddit · 2026-07-11
The author reviewed their experience after returning to OpenAI, with the core takeaway being that OpenAI's technology and product form have evolved rapidly over the past few months.
Main Takeaways
- They had previously switched to Claude / Claude Code, finding it more cost-effective for coding and automation tasks.
- However, ChatGPT 5.5 performed very stably in their daily cybersecurity work, especially for technical coding and security-related tasks.
- Still, they felt the pure chat experience remained somewhat "dumb."
Thoughts on 5.6 and the New Platform
- They noted that after the launch of 5.6, the platform experience for Codex / ChatGPT noticeably upgraded.
- The new ChatGPT Work / Projects made context and memory much better to use, feeling like "a giant brain."
- They specifically highlighted the cheaper pricing: starting at $1 for input and $6 for output, which is more affordable than the Sonnet 4.6 API.
Overall Conclusion
They suggest testing 5.6 and the new platform with real-world use cases to see if they fit your workflow, and expressed approval of the OpenAI team's product direction.
More from Models
- Gary Marcus says LLM math skills are like knowing only a car’s engine size — GaryMarcus · 2026-07-22
- OpenAI’s Codex + GPT-5.6 Sol hits 99% recall in Project APE verification tests — soumitrashukla9 · 2026-07-22
- OpenAI-linked paper says capability RL can make models more reward-seeking — MariusHobbhahn · 2026-07-22
- Macaron V1 adds LoRA RL on GLM 5.2 and claims SOTA benchmark gains — Xianbao_QIAN · 2026-07-22
- OpenAI rolls out voice in GPT-Live, but the UI obscures search and reasoning — Graham_dePenros · 2026-07-22
- Moonshot’s Kimi K3 sets a new open-weights ECI record at 156 — scaling01 · 2026-07-22