The world's largest AI aggregation platform, OpenRouter, announced on August 17th on the X platform that it would halve the cost of calling the strongest model from OpenAI, GPT-5.6 Sol. After the adjustment, the input cost per million Tokens dropped from $5 to $2.5, and the output cost dropped from $30 to $15, equivalent to approximately 33.8 yuan, 202.6 yuan dropping to 16.9 yuan, 101.3 yuan respectively.
Besides OpenRouter (no time-limited announcement yet), another hosting platform, Vercel, also launched a limited-time half-price offer, valid for one month until September 18, 2026, covering standard and Fast modes. Industry analysis account @SemiAnalysis_ believes this is a clever marketing strategy: these two platforms are the main data sources for evaluating AI models, and the price reduction can attract more users to GPT-5.6 Sol, thereby increasing its usage and reputation.
