GPT-5.5 ships with 82.7% Terminal-Bench 2.0 — and a 2x price hike that's splitting developers

OpenAI launched GPT-5.5, posting 82.7% on Terminal-Bench 2.0 (up from 75.1% on 5.4) and 73.1% on Expert-SWE, with a 1M-token context window. The company doubled pricing to $5 input / $30 output per million tokens, arguing token-efficiency gains net out the increase. ZDNET's 10-round eval scored it 93/100, docking points only for 'exuberance.' OpenAI also positioned 5.5 as a step toward a unified 'super app' merging ChatGPT, coding, and browser capabilities.