TL;DR
OpenAI has released GPT-5.2, its most capable model series yet for professional knowledge work. The new model outperforms industry professionals on 70.9% of knowledge work tasks and delivers significant improvements in coding, vision, and tool-calling capabilities.
A New Standard for Enterprise AI
OpenAI claims GPT-5.2 represents a substantial leap in AI capabilities for professional applications. According to the company, the average ChatGPT Enterprise user reports saving 40-60 minutes daily, with heavy users claiming over 10 hours weekly in productivity gains.
The model achieves what OpenAI describes as “state of the art” results across multiple benchmarks, including their GDPval evaluation spanning 44 occupations. On professional tasks like creating presentations, spreadsheets, and code, GPT-5.2 Thinking reportedly produces outputs at more than 11 times the speed and less than 1% of the cost of expert professionals.
Key Technical Improvements
GPT-5.2 introduces several notable enhancements:
- Coding performance: 55.6% on SWE-Bench Pro, a rigorous multi-language software engineering evaluation
- Reduced hallucinations: 30% fewer responses containing errors compared to GPT-5.1
- Long-context reasoning: Near 100% accuracy on 4-needle MRCR tasks out to 256,000 tokens
- Vision capabilities: Roughly halved error rates on chart reasoning and interface understanding
Major enterprise customers including Notion, Shopify, Databricks, and JetBrains have provided early testimonials supporting the performance claims.
Looking Forward
GPT-5.2 is available now to paid ChatGPT subscribers (Plus, Pro, Business, Enterprise) with API access at $1.75 per million input tokens. The release signals OpenAI’s continued push toward AI that can handle complex, multi-step professional workflows with minimal human intervention.
Source: OpenAI