AI development is moving past just answering questions and getting into systems that can think, write code, use computers and manage complex tasks step by step. OpenAI’s GPT-6.1 Sol is the latest addition to this shift, bringing performance close to its more powerful Astra model while targeting lower operating costs. For companies creating AI products this update is very important because it brings together thinking skills, with coding, computer use and professional work processes. This makes advanced AI easier to use across businesses.
Release Notes
GPT-6.1 Sol is positioned as a lower-cost alternative for workloads that require substantial reasoning and technical capability. OpenAI describes it as delivering near-Astra performance across complex coding, computer use, and professional tasks.
The model is available through the OpenAI API and is also rolling out across Work and Codex for supported users. Its API pricing is set at $2 per million input tokens and $10 per million output tokens, while the model supports a context window of up to 1.05 million tokens.
Standout Features
- Advanced Reasoning: GPT-6.1 Sol supports multiple reasoning levels, allowing developers to balance response quality, processing effort, and application requirements.
- Complex Coding: The model is designed for software development tasks that involve understanding larger codebases, solving technical problems, and working through multi-step implementation requirements.
- Agentic Workflows: Built-in support for tools such as web search, file search, code execution, computer use, and MCP enables applications to move beyond text generation.
- Large Context Handling: A 1.05-million-token context window allows the model to process substantial amounts of project information within a single workflow.
- Structured AI Integration: Function calling and structured outputs make it easier to connect model responses with business applications and automated processes.
Real-World Applications
For companies GPT-6.1 Sol can help with AI coding helpers, company information systems, tasks that require a lot of documents, software automation, customer service robots, research tools and internal work tools. The mix of features and lower costs can also make using AI a lot more practical especially when tasks need repeated thinking or using tools.
For teams working on AI the model gives another choice when building systems that need to balance how well they work, how big they can get and how much they cost to run.
Conclusion
In the end GPT-6.1 Sol shows a shift toward AI models designed for real-life tasks that take multiple steps instead of just handling single questions. For companies that are thinking about their AI product or improving an old system, picking the correct model is now a key part of the design.
Want to create or add AI to your company? Work with ScalaCode to change new AI features into solutions that can grow and work in situations.
