OpenAI
GPT-6.1 Sol: OpenAI bets on speed with a model five times cheaper
🧒 Discovery
OpenAI made an AI that answers almost as well as before, but much faster and much cheaper!
📰 Essential
OpenAI launched GPT-6.1 Sol and a new Ultrafast tier that generates 300 tokens per second, offering performance close to Astra at one-fifth of the price.
🔬 Expert
The Ultrafast tier targets use cases where latency matters as much as quality, such as live customer support or real-time financial analysis, with a claimed throughput of 300 tokens per second. By pricing this model five times cheaper than Astra for comparable performance, OpenAI is clearly aiming at high-volume workloads where rivals currently undercut it on cost.