OpenAI has introduced GPT-6.2 Astra and Sol, new models designed for better professional outputs, coding, and computer use. Available via ChatGPT Work, Codex, and the API, these models offer higher efficiency at a lower cost for business users and developers.
OpenAI just unveiled GPT-6.2 Astra and Sol, a pair of models built to move AI from a chat box into a functional workspace. These models focus on professional-grade writing, advanced coding, and a new ability called "computer use."
If you use ChatGPT for work or build apps with the API, these updates are about making the AI more capable and cheaper to run. We will break down what this means for your monthly bill and your daily tasks.
You will learn exactly where these models are available and how to start using them to automate the boring parts of your professional life.
Key takeaways
- GPT-6.2 Astra and Sol improve coding and professional outputs.
- New "computer use" capabilities allow the AI to interact with software.
- API and ChatGPT Work costs are now lower.
- Access is provided through Codex, the API, and ChatGPT Work.
What GPT-6.2 Astra and Sol bring to the table
OpenAI's new models focus on making AI a more active participant in professional workflows. According to OpenAI, GPT-6.2 Astra and Sol provide better professional outputs, coding, and computer use [openai.com].
Instead of just generating text, these models are built to handle complex technical tasks. This includes writing more reliable code and interacting with computer interfaces to perform actions that previously required a human to click through software. For users of Codex and the API, this means the AI can act more like a digital assistant that can actually execute steps in a program.
Because these models are designed for efficiency, they are also cheaper to use. This is a significant shift for anyone running high-volume tasks through the OpenAI API, as it allows for more complex automation without a proportional increase in your monthly bill. You can now focus on the result rather than the cost of the tokens used to get there.
At a glance
| Detail | What we know |
|---|---|
| Who | OpenAI |
| What | GPT-6.2 Astra and Sol |
| Where | ChatGPT Work, Codex, API |
| Primary Goal | Better professional outputs, coding, and computer use |
| Cost | Lower cost |
| Efficiency | Improved |
| Release Date | Not announced |
| Regional Limits | Not announced |
Why OpenAI lowered costs for Astra and Sol
OpenAI is targeting the professional market by making these tools more efficient and affordable. The new models are available at a lower cost across ChatGPT Work, Codex, and the API [openai.com].
For small business owners and content creators, the lower cost is the most immediate benefit. If you use AI to handle repetitive tasks like data entry or social media scheduling, the efficiency of these models means you can do more with the same budget. If you are managing costs for a growing team, you can use our Token and cost calculator to plan your transition to these new models once pricing is finalized.
This move suggests OpenAI wants to move away from the "experimental" phase of AI and into a phase where AI is a standard utility in the office. By lowering the price, they make it easier for a company to let the AI handle the first draft of every professional document or the first pass of every piece of code.
AI primarily generated text or code snippets for you to copy.
AI can assist with computer use and professional outputs more efficiently.
You can automate technical tasks at a lower cost.
- Exact pricing per token for Astra and Sol
- Specific functional differences between the Astra and Sol models
- Regional rollout dates for ChatGPT Work
How GPT-6.2 Astra and Sol handle coding
OpenAI has integrated these models into Codex and the API to specifically improve technical outputs. This means developers can expect better coding assistance and more reliable scripts [openai.com].
Developers will find these updates particularly useful because the models are tuned for professional-grade coding. This reduces the time spent debugging AI-generated scripts and allows for more complex architecture suggestions. As you integrate these models, you might find it helpful to use our AI model picker to compare how Astra and Sol perform against your existing setup.
Combining better coding with lower costs means you can build more robust AI-powered features into your own apps without blowing your budget. You can also check our AI news section for updates on specific release dates for your region as these models roll out to more developers.
Hype vs reality
These models will replace all professional coders.
OpenAI describes them as providing "better" coding and professional outputs, not total replacement.
How it compares
| Feature | GPT-6.2 Astra/Sol | Previous Models |
|---|---|---|
| Focus | Professional/Coding/Computer Use | General Conversation |
| Cost | Lower | Standard |
| Interface | Work/API/Codex | Chat/API |
What it means for you
If you run a small business
You can automate more software tasks with lower overhead.
Do this: List repetitive software steps you do daily.
If you are a developer
You get cheaper API calls and better code assistance.
Do this: Audit your current API usage to prepare for a switch.
If you are a professional writer
Higher quality professional outputs reduce your editing time.
Do this: Test the model with a complex professional report.
Who should care (and who can skip it)
Worth your time if you…
- You use ChatGPT Work for your job
- You build apps with the OpenAI API
- You need high-end coding assistance
You can skip it if you…
- You only use AI for casual conversations
- You do not use professional AI tools
How to get ready
- Check your subscription Ensure you have access to ChatGPT Work or Codex to receive the update.
- Review your API budget Check your current spending to see where lower costs will help most.
- Identify automation targets List software tasks you currently do manually that could benefit from computer use.
Tests professional output and software-oriented tasks.
Act as a professional [JOB TITLE]. Using your computer use capabilities, help me [SPECIFIC TASK, e.g., format this data for a spreadsheet] and ensure it follows [SPECIFIC PROFESSIONAL STANDARD].
Be explicit about the professional standard you expect, such as specifying a particular industry style guide or coding standard.
Expecting the AI to have full, unsupervised control of your computer. Always review computer use actions before they are finalized.
What to watch next
- Detailed Pricing To see the exact savings for your specific API volume.
- Sol vs Astra docs To know which specific model to pick for different tasks.
Our take
GPT-6.2 Astra and Sol are a pragmatic update for the professional crowd. The lower cost is the biggest win, while computer use is the most intriguing feature. The main caveat is the lack of detail on how computer use actually functions in practice.
Identify one repetitive software task you do daily that you want the AI to handle.
Your next step
- GuideAI models guide
Understand where these professional models fit in the wider landscape.
- By professionPrompts for your job
Find specific professional prompts to test the new capabilities.
- Free toolVideo to Prompt
Questions people ask
What is the difference between Astra and Sol?
OpenAI has not yet announced specific differences between the Astra and Sol models. They are presented together as part of the GPT-6.2 update for professional use.
Will GPT-6.2 be cheaper to use?
Yes, OpenAI states that the GPT-6.2 Astra and Sol models are designed for better efficiency at a lower cost across ChatGPT Work and the API.
What does 'computer use' mean in GPT-6.2?
Computer use refers to the model's ability to interact with software and computer interfaces to complete tasks rather than just providing text.
Where can I access these new models?
OpenAI is making GPT-6.2 Astra and Sol available through ChatGPT Work, Codex, and the API.
Sources checked . We update this story when details change.
Written with AI assistance from the sources above and checked against them before publishing. Spotted an error? Tell us and we will fix it.