Skip to main content

OpenAI Releases GPT-6.1 Sol at One-Fifth Astra Price

OpenAI launches GPT-6.1 Sol via API, ChatGPT Work, and Codex at 1/5th the token price of Astra, with an Ultrafast mode coming soon.

AI-written
Inewgen
01 Oct 2026Source: MarkTechPost2 min read (0 views)
Share
OpenAI Releases GPT-6.1 Sol at One-Fifth Astra Price

Stock photo for illustration only, not from the actual event

Font size
  • OpenAI introduces GPT-6.1 Sol, now live in the API, ChatGPT Work, and Codex
  • Priced at one-fifth of GPT-6 Astra, costing $2 per million input tokens
  • An Ultrafast option is coming to Codex soon, promising up to 8x faster generation

OpenAI has officially rolled out its newest hosted model, GPT-6.1 Sol, available starting today within the OpenAI API under the identifier gpt-6.1-sol. The model is also integrated into ChatGPT Work and Codex, bringing near-Astra coding and computer use capabilities at a significantly reduced price point.

This release establishes three distinct tiers within OpenAI's GPT-6 lineup. The premium tier, GPT-6 Astra, sits at the top, followed by the newly introduced mid-tier GPT-6.1 Sol, and the lightweight GPT-6 Luna designed for high-efficiency, low-cost operations.

$2GPT-6.1 Sol input cost per million tokens
1/5Price ratio compared to Astra tier
8xSpeed boost for the upcoming Ultrafast mode

According to verified first-party API list prices, GPT-6 Astra costs $10 for input, $50 for output, and $1 for cached input per million tokens. In comparison, GPT-6.1 Sol reduces costs to $2, $10, and $0.10, while the budget-friendly GPT-6 Luna drops down to $0.10, $0.50, and $0.01 respectively.

notebook computer office desk workspace

Stock photo for illustration only, not from the actual event

The cached pricing structure plays a critical role for AI agents, which repeatedly resend the same system prompts, tool schemas, and history logs across execution steps. Cached reads for GPT-6.1 Sol now cost only 5% of the uncached input rate, improving upon the 10% rate seen previously and optimizing long-running agent workflows.

Never miss the latest news?

Subscribe to get news summaries by email - not often enough to be annoying.

โฆษณา

By introducing a mid-tier model at one-fifth of the flagship price and optimizing cache rates, OpenAI is directly targeting the economic bottlenecks of multi-step AI agents. Lowering token expenditure makes continuous agentic loops financially viable for independent developers and enterprises alike.

Looking ahead, OpenAI also plans to introduce a GPT-6.1 Sol Ultrafast option inside Codex within the coming days, targeting up to 8x faster token generation speeds compared to standard operating rates.

Source: MarkTechPost

Comments

Leave a Comment
0/2000

Found something wrong in this article? Report an issue with this article