**OpenAI’s GPT‑5.6 Launch Is a Direct Response to Anthropic’s Momentum**
In a flurry of announcements, OpenAI has rolled out a new set of models and tools that appear calibrated to counter the buzz surrounding Anthropic’s recent releases. The moves include three new GPT‑5.6 variants, clearer benchmarking comparisons, and an agentic desktop productivity feature that directly parallels Anthropic’s Claude Cowork.
**Three New Models: Sol, Terra, and Luna**
OpenAI’s new models follow a celestial naming scheme that maps closely to GPT‑5.5 tiers:
– **Sol** – Flagship model designed to compete with Anthropic’s Fable 5, roughly equivalent to GPT‑5.5’s “thinking” mode.
– **Terra** – The default, mainstream GPT‑5.6 model, aligned with GPT‑5.5’s standard tier.
– **Luna** – A lighter, faster option, comparable to GPT‑5.5’s instant model.
**Benchmarks Target Claude Performance**
OpenAI highlighted several benchmark results positioning GPT‑5.6 Sol against Claude Fable 5:
– **Agents’ Last Exam**: Sol scored 53.6, beating Claude Fable 5 by 13.1 points.
– **Artificial Analysis Intelligence Index**: Sol came within one point of Fable 5 while completing tasks 61% faster.
– **Artificial Analysis Coding Agent Index**: Sol reached 80, 2.8 points above Fable 5.
OpenAI also emphasized efficiency, claiming Sol can outperform Fable 5 at roughly one‑quarter the cost at medium reasoning levels. In internal tests, GPT‑5.6 showed stronger agentic code‑review performance than GPT‑5.5, with higher accuracy and lower latency.
**ChatGPT Work: OpenAI’s Answer to Claude Cowork**
Alongside the models, OpenAI launched **ChatGPT Work**, an agentic desktop tool that can control your computer, manage files, build spreadsheets, create presentations, and run complex, multi‑step projects. Like Claude Cowork, it integrates with external apps and services, including Slack, Teams, Google Drive, email, and CRMs.
In a dramatic demo, an OpenAI employee used ChatGPT Work to completely reorganize Apple Notes. While powerful, such deep system access also underscores ongoing trust and safety considerations for users delegating sensitive workflows to AI.
**Availability and Rollout**
ChatGPT Work is initially available to Pro, Enterprise, and Education users on web and mobile, with Plus and Business tiers to follow. OpenAI is consolidating its separate ChatGPT and Codex desktop apps into a single application, mirroring Anthropic’s strategy of bundling chat, coding, and cowork capabilities.
**A Clear Catch‑Up Strategy**
Taken together, the announcements suggest OpenAI is accelerating its response to Anthropic’s growing influence. By pairing high‑performance, benchmark‑leading models with an agentic productivity tool, OpenAI is positioning GPT‑5.6 as a more capable, efficient, and commercially attractive alternative.
*Original article source: [ZDNET – OpenAI’s briefing on the new models](https://www.zdnet.com/article/openai-gpt-5-6-anthropic-claude-catchup/)*
