NewsBite analysis · AI, Data · New Product, Software Release
The Decoder
GPT-5.6 Sol goes 14x faster as OpenAI launches Ultrafast mode powered by Cerebras
OpenAI has announced a preview of its 'Ultrafast' mode for GPT-5.6 Sol, which allows output speeds of up to 750 tokens per second. This mode aims to combine the speed of smaller models with the full reasoning capabilities of larger models. Applications range from incident response, where engineers can analyze data during outages, to finance and e-commerce, where fast decision-making is critical. The model's real-time capabilities are expected to transform workflows in research and customer support by enabling immediate insights and adjustments.
MonitorConfidence: HighShort term horizon
Published 14 Aug 2026, 16:10 UTC2 min read
Visit original article ↗Lead image — The Decoder
Summary
TL;DR
- Ultrafast mode delivers up to 750 tokens per second.
- Combines speed and reasoning capabilities of large models.
- Applicable in incident response, finance, and customer support.
- Transforms overnight batch jobs into interactive sessions.
- Reflects a tiered pricing model similar to cloud services.
Interesting facts
- OpenAI currently offers a 'Fast Mode' priced at double for lower speeds.
- This new tier aims to capitalize on faster inference across industries.
Coverage gaps