> ## Content Index
> Fetch the complete content index at: https://nextwith.ai/llms.txt
> Use this file to discover other available public pages before exploring further.

# OpenAI launches GPT-6.1 Sol at one-fifth GPT-6 Astra’s token price
- URL: https://nextwith.ai/openai-launches-gpt-6-1-sol-at-one-fifth-gpt-6-astras-token-price/
- Published: 2026-09-30T09:28:50.000Z
- Updated: 2026-09-30T09:28:50.000Z
- Description: OpenAI has released GPT-6.1 Sol, a lower-cost model that the company says nears GPT-6 Astra on coding and agentic work. The published pricing tables show standard text tokens at one-fifth of Astra’s rates, with important deployment limits.
- Author: NextWith.ai Editorial Desk
- Tags: AI Models, News

## OpenAI has added a cheaper model to its top-tier lineup

OpenAI introduced GPT-6.1 Sol at DevDay, and TechCrunch reported that the company made it available that day to Plus, Pro, Business, Enterprise and Edu users in ChatGPT Work and Codex, while not yet offering it in Chat. That is the confirmed development here: a new model, new pricing, and immediate access in OpenAI’s agent-oriented products, not just a lab preview.

On OpenAI’s [GPT-6.1 Sol model page](https://developers.openai.com/api/docs/models/gpt-6.1-sol?ref=nextwith.ai), the company describes the model as a lower-cost option for complex coding, computer use and professional work, with “near-Astra performance” for those tasks. TechCrunch reported that OpenAI framed the model as nearly matching GPT-6 Astra on agentic coding, computer use and professional work. Those are OpenAI’s own claims, so they are useful for understanding the release but not independent proof of performance.

The practical significance is that Sol is not positioned as a stripped-down utility model. OpenAI lists the same 1,050,000-token context window and 128,000-token output limit that appear on the [GPT-6 Astra model page](https://developers.openai.com/api/docs/models/gpt-6-astra?ref=nextwith.ai). The difference is the price, not the basic scale of the model’s working window.

## The pricing math is the real headline

OpenAI’s published pricing makes the comparison unusually direct. Sol is listed at $2 per million input tokens and $10 per million output tokens. Astra is listed at $10 per million input tokens and $50 per million output tokens. On standard text pricing, that is exactly one-fifth of Astra’s rate on both input and output. The company also lists cheaper cached-input pricing for both models, but the same 5x gap remains in the standard figures that headline the release.

That matters because many AI budgets are dominated by token volume, not one-off API calls. Long coding sessions, multi-step agent runs, document review, and computer-use workflows can burn through tokens quickly. If a team is paying for repeated reasoning, tool calls and long context, a model that costs one-fifth as much can change which tasks are economical to automate and which still need a higher-end model.

Here the useful distinction is between OpenAI’s claim and the operational implication. OpenAI is claiming Sol gets close to Astra on difficult work; the operational implication is that, if that claim holds well enough for a given use case, teams may be able to route routine or moderate-complexity jobs to Sol and reserve Astra for the cases where accuracy, robustness or judgment matters more than cost.

## How it is meant to be used, and where the boundary is

The documentation also shows that OpenAI is treating Sol as a tool-heavy model rather than a bare chat endpoint. The model page says to use the Responses API for tool calling, while Chat Completions works without tool calling. It also says Sol supports web search, file search, image generation, code interpreter, hosted shell, apply patch, skills, computer use, MCP and tool search when used through the Responses API. For product teams, that is a signal that Sol is intended to sit inside agent workflows, not just answer prompts.

There are also practical constraints. OpenAI says Sol supports US and EU data residency, but fast mode is unavailable with EU data residency. The pricing notes further say fast mode costs 2x standard, batch and flex are 50% lower than standard, regional processing can add a 10% premium where available, and prompts over 272K input tokens are charged at higher rates. In other words, the one-fifth headline is real, but it only describes the standard case; deployment details can push the actual bill up or down.

The model page also lists reasoning.effort settings from low to max, with none and minimal not supported. That suggests OpenAI expects developers to tune the model’s reasoning budget to the task. For teams building internal copilots, customer-support agents or document-processing systems, that is important because it creates a decision boundary: the cheapest setup may not be the most reliable one, and the right configuration may depend on whether the job is draft generation, tool use, or a workflow that needs strict adherence to constraints.

OpenAI’s release therefore looks less like a simple price cut and more like a rebalancing of its model ladder. Sol gives developers a cheaper path into the same high-context, tool-enabled workflow stack as Astra, but the company’s own documentation still leaves the burden on buyers to decide where the quality trade-off is acceptable. Before moving production workloads from Astra to Sol, compare your own coding, tool-use, and factuality runs against the published price gap so you can decide whether the lower cost outweighs any quality risk.