GPT-6.1 Sol: What’s New?
OpenAI has announced that the API pricing for GPT-6.1 Sol remains consistent with its predecessor, GPT-6 Sol. This means businesses will continue to pay $2 for input and $10 for output per million tokens. However, there’s a significant change in the cost of cache reading, which has been slashed from $0.20 to $0.10.
Immediate Financial Impact
- Cost Reduction: The reduction in cache reading costs offers a potential saving of up to 21% for businesses heavily reliant on cache. However, for a typical agent session, the savings are only around 4%.
- Usage Increase: Bain has observed that while the token price has decreased, the overall usage has increased, potentially offsetting the savings.
Market Dynamics
- AI API Market: The reduction in cache reading costs could make AI solutions more accessible, potentially expanding the market for AI APIs.
- OpenAI’s Role: OpenAI remains a pivotal player, especially with its recent agreement with the U.S. military, which could influence market dynamics.
Challenges Ahead
- Balancing Costs and Usage: The main challenge for SMEs will be to balance the reduced costs with the increased usage. The danger lies in the possibility that the cost reduction might not be enough to attract new users if usage continues to rise.
Conclusion
For SMEs, the changes in GPT-6.1 Sol’s pricing structure present both opportunities and challenges. While there are potential savings, the real impact will depend on how businesses manage their usage and adapt to the evolving AI API market.