OpenAI Releases GPT-6.1 Sol to Slash Intelligence Costs
OpenAI has introduced GPT-6.1 Sol, a new model designed to deliver high-tier intelligence comparable to Astra at just one-fifth of the cost, intensifying an industry-wide price war.
During its DevDay 2026 keynote on September 29, 2026, OpenAI announced the release of GPT-6.1 Sol. The new model aims to provide developers with highly advanced capabilities at a fraction of the traditional cost. According to analysis by technologist Simon Willison, the model delivers "near-Astra intelligence for a fifth of the price," marking a significant shift in the economics of frontier artificial intelligence.
The launch of GPT-6.1 Sol comes amid a rapidly escalating price war among major AI providers. Just days prior, on September 22, 2026, the industry saw major shifts with the discussion of Claude Opus 5.5 alongside OpenAI's own GPT-6 Sol and GPT-6 Luna. The introduction of the GPT-6.1 Sol variant represents a rapid iteration for OpenAI as it seeks to maintain its competitive edge against rivals like Anthropic while making powerful models financially viable for enterprise scale.
For practitioners, the release of GPT-6.1 Sol changes the calculus of deploying advanced LLMs. Visualizations of the model's capabilities, referred to as pelican charts, indicate that GPT-6.1 Sol does not deviate significantly from the performance benchmarks of the standard GPT-6 family. This means developers can access top-tier reasoning and generation features without paying the premium prices previously associated with frontier models.
Ultimately, the arrival of GPT-6.1 Sol highlights a broader trend in 2026 where frontier AI companies are focusing as much on cost reduction and efficiency as they are on raw capability gains. By offering near-Astra intelligence at an eighty percent discount, OpenAI is forcing the market to adapt, making high-performance AI integration far more accessible for budget-conscious developers and enterprises alike.
This is our own summary of reporting by Simon Willison



