OpenAI has introduced an "Ultrafast Mode" for its GPT-5.6 Sol model, enabling speeds up to 14 times faster than previous iterations. This new service tier, available through the OpenAI API and powered by Cerebras, can deliver up to 750 output tokens per second. Benchmarks show GPT-5.6 Sol in Ultrafast mode significantly outperforming models like Claude Fable-5 and Opus 4.8 in speed, particularly on challenging benchmarks such as Humanity's Last Exam, while maintaining comparable accuracy. AI
IMPACT Accelerates time-sensitive applications and new agent-based workflows by resolving the speed-intelligence tradeoff.
RANK_REASON Frontier-lab model release with system card. [lever_c_demoted from frontier_release: ic=2 ai=1.0]
- Cerebras
- GPT 5.6 "Sol"
- OpenAI
- Ultrafast Mode
- Artificial Analysis GPT-5.6 Sol
- Claude Fable-5
- Humanity's Last Exam
- OpenAI API
- Opus 4.8
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
