What happened
OpenAI launched GPT-6.1 Sol, a new AI model positioned as a cost-effective alternative to its flagship GPT-6 Astra. The model is available to Plus, Pro, Business, Enterprise, and Edu users via ChatGPT Work and Codex, as well as through the OpenAI API. OpenAI reports that GPT-6.1 Sol costs roughly one-fifth as much as Astra while matching or exceeding performance on several key benchmarks, including software engineering and computer use tasks.
OpenAI has introduced GPT-6.1 Sol, a new model designed to offer performance close to its more powerful GPT-6 Astra at a significantly lower cost. According to the report, the model improves on the previous GPT-6 Sol in coding, computer use, and professional work tasks. The pricing structure is a key differentiator, with OpenAI stating that GPT-6.1 Sol costs roughly one-fifth as much as Astra based on standard API input and output prices.
Specific pricing details provided by OpenAI include $2 USD per million input tokens, $0.10 USD per million cached input tokens, and $10 USD per million output tokens. The company notes that cached inputs are 95 percent cheaper than regular inputs and half the cost of cached inputs for the previous GPT-6 Sol model. This pricing strategy aims to make advanced AI capabilities more accessible to developers and businesses with varying budgets.
results published by OpenAI show GPT-6.1 Sol matching GPT-6 Astra on DeepSWE v1.1, a software engineering benchmark, while scoring 6.4 percentage points higher than GPT-6 Sol. In professional PDF tasks measured by GDP.pdf, the new model outperformed Opus 5.5 with fallbacks at less than half the cost per task. On AutomationBench, GPT-6.1 Sol finished 2.2 percentage points ahead of Opus 5.5 at roughly one-third of the cost.
The model also shows improvements in computer use and factual accuracy. GPT-6.1 Sol scored seven percentage points higher than GPT-6 Sol on OSWorld 2.0 and came within 2.1 points of GPT-6 Astra. On Terminal-Bench Science, it more than doubled the score of GPT-6 Sol. Additionally, the rate of factual errors decreased from 11.4 percent in GPT-6 Sol to 7.7 percent in GPT-6.1 Sol at low reasoning effort.
Availability is currently limited to Plus, Pro, Business, Enterprise, and Edu users through ChatGPT Work and Codex, with API access under the name gpt-6.1-sol. It is not yet available in regular ChatGPT conversations. OpenAI also announced that a GPT-6.1 Sol Ultrafast version is coming to Codex in the next few days, offering up to eight times faster generation.
Source details: iphoneincanada.ca ↗
Why it matters
The release of GPT-6.1 Sol represents a significant shift in the AI market by offering near-flagship performance at a substantially lower price point. By reducing the cost per and per task, OpenAI makes advanced AI capabilities more accessible to a broader range of developers and enterprises. This move intensifies competition in the AI model market, potentially forcing other providers to adjust their pricing strategies or improve their mid-tier models to remain competitive. For users, this means more affordable access to high-performance AI for coding, automation, and professional tasks.
The launch of GPT-6.1 Sol is significant because it bridges the gap between high-performance flagship models and cost-effective mid-tier options. By offering near-Astra performance at a fraction of the cost, OpenAI is likely to drive adoption among developers and enterprises who previously found flagship models too expensive for routine use.
This move has practical implications for AI integration in business workflows. The reduced cost per task, particularly in automation and coding, makes it feasible for smaller organizations to implement advanced AI solutions. The improvement in factual accuracy and computer use capabilities further enhances the model's utility for professional applications.
The competitive landscape is also affected. With GPT-6.1 Sol outperforming or matching competitors like Opus 5.5 on key benchmarks at lower costs, other AI providers may need to reassess their pricing and product strategies. This could lead to a broader trend of cost reduction and performance improvement across the AI industry.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
In AI, what are a model's "parameters"?
What to watch next
Monitor the performance of GPT-6.1 Sol in real-world applications compared to its claims. Watch for the release of the GPT-6.1 Sol Ultrafast version in Codex, which promises faster generation. Additionally, observe how competitors like Anthropic and Google respond to this pricing and performance shift in the mid-to-high tier AI model market.
The real-world performance of GPT-6.1 Sol in diverse applications will be a key indicator of its success. While results are promising, actual user experiences and feedback will provide a more comprehensive picture of the model's capabilities and limitations.
The upcoming release of the GPT-6.1 Sol Ultrafast version in Codex is worth monitoring. Faster generation could further enhance the model's appeal for time-sensitive tasks and real-time applications, potentially setting a new standard for AI speed and efficiency.
Competitor responses to the launch of GPT-6.1 Sol will be crucial. Watch for announcements from Anthropic, Google, and other AI providers regarding new models or pricing adjustments that aim to counter OpenAI's latest offering. The market's reaction to this cost-performance shift will shape the future of AI model development and deployment.