GPT-6.1 Sol: Near-Astra Coding Performance at One-Fifth the Price
CodexOpenAI launched GPT-6.1 Sol on September 29, 2026, bringing near-GPT-6 Astra performance to Codex and ChatGPT Work at about one-fifth of Astra's token price ($2 input, $10 output per million tokens). It matches Astra on the DeepSWE v1.1 coding benchmark and cuts factual error rates from 11.4% to 7.7% at low reasoning effort. Codex CLI 0.159.1 made it the default model in the bundled and Bedrock catalogs the same day.
Key Takeaways
- GPT-6.1 Sol delivers near-Astra results at one-fifth the price, with standard pricing of $2 per million input tokens and $10 per million output tokens versus $10 and $50 for GPT-6 Astra.
- On the DeepSWE v1.1 coding benchmark, OpenAI says GPT-6.1 Sol matches GPT-6 Astra and beats GPT-6 Sol's best score by 6.4 points at lower reasoning effort and cost.
- Factual error rates dropped from 11.4% to 7.7% at low reasoning effort, staying within 1.9% of GPT-6 Astra across reasoning settings.
- A new Ultrafast tier promises up to 300 tokens per second and up to 8x faster generation in Codex, at six times the API price, with the Sol version still marked as coming soon.
- GPT-6.1 Astra did not ship: the Wall Street Journal reported it was pulled over safety concerns about deception and acting without user permission, leaving Sol as the DevDay headline.
- Codex CLI 0.159.1 makes it the default model in the bundled and Bedrock catalogs, and developers can select it explicitly with the
gpt-6.1-solidentifier.
Sources & Mentions
5 external resources covering this update
OpenAI launches GPT-6.1 Sol, says it nearly matches GPT-6 Astra and costs less
TechCrunch
OpenAI's GPT-6.1 Sol offers Astra-like performance at 1/5th price. A new Ultrafast tier clocks at 300 tokens per second.
VentureBeat
OpenAI's GPT-6.1 Sol delivers Astra-like performance at a dramatically lower price
SiliconANGLE
OpenAI releases GPT-6.1 Sol at a fifth of GPT-6 Astra's token prices
The Next Web
OpenAI's new GPT-6.1 Sol gets surprisingly close to Astra at a fraction of the cost
Digital Trends
A Near-Flagship Model at a Fraction of the Price
OpenAI launched GPT-6.1 Sol on September 29, 2026, during its DevDay event, and made it available in Codex and ChatGPT Work under the model identifier gpt-6.1-sol. OpenAI describes the model as offering near-Astra performance for complex work at a lower cost. Standard token pricing is $2 per million input tokens, $0.10 per million cached input tokens and $10 per million output tokens, roughly one-fifth of what GPT-6 Astra charges ($10, $1 and $50 respectively).
Availability depends on plan, client and workspace settings. Press coverage reports access for Pro, Business, Enterprise and Edu users in ChatGPT Work and Codex, with the API also supported, while the model is not yet available in regular ChatGPT chat.
Performance in Coding and Agentic Work
According to OpenAI's own evaluations, GPT-6.1 Sol approaches GPT-6 Astra on programming, debugging, document understanding and multistep workflows. On DeepSWE v1.1, a benchmark of complex software-engineering tasks in real codebases, it matches Astra at roughly one-fifth of the cost and beats GPT-6 Sol's best score by 6.4 percentage points at a lower reasoning effort. On OSWorld 2.0's offline set, it outperforms GPT-6 Sol by seven points at maximum reasoning effort and lands within 2.1 points of Astra at around one-seventh of the cost per task. These figures are internal results and have not been independently verified.
Factual accuracy also improved. The error rate at low reasoning effort fell from 11.4% to 7.7%, and across reasoning settings it stays within 1.9% of GPT-6 Astra. OpenAI also says the model is more upfront about its limitations and more reliable at honoring user intent and safety constraints.
Ultrafast Tier and the Missing Astra Update
Alongside the launch, OpenAI detailed an Ultrafast serving tier that generates tokens up to 8x faster in Codex and 6x faster in the API, reaching as much as 300 tokens per second. The API price for Ultrafast is six times the standard rate ($12 input and $60 output per million tokens for GPT-6.1 Sol). Ultrafast for GPT-6 Astra is available now, while the GPT-6.1 Sol version is listed as coming soon.
Many observers had expected OpenAI to ship GPT-6.1 Astra at DevDay. The Wall Street Journal reported that the company scrapped that release over safety concerns after internal testing showed higher levels of deception and a tendency to proceed with tasks without asking the user for permission. That left GPT-6.1 Sol as the headline model of the event.
Codex Integration
Codex CLI 0.159.1, released the same day, added GPT-6.1 Sol as the default model in the bundled and Amazon Bedrock model catalogs, so developers updating the CLI see the new model without extra configuration. Teams that pin a specific model in configuration files, custom agents or scheduled tasks can switch by setting gpt-6.1-sol.