OpenAI has introduced GPT-6.1 Sol, an upgraded version of GPT-6 Sol that the company says approaches the performance of its flagship GPT-6 Astra across coding, computer use and professional workflows. The new model is priced at one-fifth of GPT-6 Astra’s standard input and output token rates.
Alongside GPT-6.1 Sol, OpenAI has introduced Ultrafast, a premium inference tier designed to provide significantly faster token generation. The new tier can deliver up to 8X faster token generation in Codex and up to 6X faster generation through the API, reaching speeds of as much as 300 tokens per second.
The announcements were reported by VentureBeat on September 29, 2026, as part of OpenAI’s broader product and developer updates.
GPT-6.1 Sol Retains the Existing Sol Pricing
GPT-6.1 Sol is priced at $2 per million input tokens, $0.10 per million cached input tokens and $10 per million output tokens.
The uncached input and output prices remain the same as GPT-6 Sol, while the cached input price has been reduced from $0.20 to $0.10 per million tokens, representing a 50% reduction.
The pricing difference is larger when compared with GPT-6 Astra. OpenAI currently charges $10 per million input tokens, $1 per million cached input tokens and $50 per million output tokens for GPT-6 Astra.
As a result, GPT-6.1 Sol costs one-fifth as much as Astra for standard uncached input and output tokens. Its cached input price is one-tenth of Astra’s rate.
OpenAI’s developer documentation also lists GPT-6.1 Sol at $2 per million input tokens and $10 per million output tokens, with cached input priced at $0.10 per million tokens.
For comparison, GPT-5.6 Sol’s current promotional pricing is $4 per million input tokens and $20 per million output tokens. OpenAI says that promotional pricing is available at least through November 21, 2026.
GPT-6.1 Sol Moves Closer to GPT-6 Astra
OpenAI says GPT-6.1 Sol delivers capabilities comparable to GPT-6 Astra while offering a lower cost for complex coding, computer use and professional work.
According to the VentureBeat report, OpenAI’s own evaluations show GPT-6.1 Sol narrowing the performance gap with Astra across several areas.
On DeepSWE v1.1, which evaluates long-running software engineering work in real codebases, OpenAI says GPT-6.1 Sol matches Astra at roughly one-fifth the cost and improves on GPT-6 Sol’s best result by 6.4 percentage points.
On GDP.pdf, a benchmark covering professional documents with charts, tables, diagrams and fine-print information, OpenAI reports that GPT-6.1 Sol scores above Anthropic’s Claude Opus 5.5 with fallbacks while costing less than half as much per task. The company also says the model approaches Astra at roughly one-fifth of its cost.
OpenAI reported similar results on AutomationBench, which tests end-to-end tasks across 47 tools covering areas such as sales, marketing, operations, support, finance and HR. The company says GPT-6.1 Sol beats Opus 5.5 by 2.2 percentage points at medium reasoning effort while costing roughly one-third as much. It also improves by 4.8 percentage points over GPT-6 Sol at the same setting.
For computer-use tasks, OpenAI reports that GPT-6.1 Sol comes within 2.1 percentage points of Astra on the OSWorld 2.0 offline set at maximum reasoning effort, while costing roughly one-seventh as much per task.
These evaluations were conducted by OpenAI, rather than by an independent testing organisation, and the company notes that its research and API evaluations can differ from production ChatGPT performance.
OpenAI Introduces Ultrafast Inference
OpenAI’s second major announcement is Ultrafast, a premium inference tier focused on reducing response latency.
Ultrafast can provide up to 8X faster generation in Codex and up to 6X faster generation through the API, with output speeds reaching as much as 300 tokens per second. OpenAI describes it as its fastest API service tier.
The higher speed comes at a higher price. API usage under Ultrafast costs six times the corresponding model’s standard rate.
Ultrafast is immediately available for GPT-6 Astra, while the VentureBeat report says a GPT-6.1 Sol version is expected soon.
OpenAI’s pricing documentation lists standard GPT-6 Astra at $10 per million input tokens, $1 per million cached input tokens and $50 per million output tokens. The published Ultrafast pricing for GPT-6 Astra is $60 per million input tokens, $6 per million cached input tokens and $300 per million output tokens.
The pricing structure gives developers different options depending on whether cost or response speed is the primary consideration. Standard inference can be used for workloads where latency is less important, while Ultrafast is aimed at applications where faster responses can be valuable.
Broader Enterprise AI Applications
The combination of GPT-6.1 Sol and Ultrafast expands the range of price and performance options available to developers building AI-powered applications.
GPT-6.1 Sol is positioned for complex coding, computer-use and professional workflows where organisations need advanced model capabilities without paying Astra-level standard token prices. OpenAI’s model documentation describes it as offering near-Astra performance at a lower cost.
Ultrafast addresses a different requirement by allowing developers to pay a premium for faster inference. The VentureBeat report identifies interactive coding, customer support, financial analysis, incident response and other human-in-the-loop applications as potential areas where response speed can be important.
OpenAI has previously experimented with high-speed inference through a preview of GPT-5.6 Sol Ultrafast on Cerebras hardware, which reached as much as 750 output tokens per second and up to 14X Standard speed for a limited group of customers.
The latest rollout expands the approach into a broader service tier while bringing the newer GPT-6 models into the company’s current developer platform.
Availability
GPT-6.1 Sol is available through the API under the model name gpt-6.1-sol. OpenAI’s documentation says the model supports coding, computer use, web search, file search, image generation and other tools through the Responses API.
The model is also positioned for use across ChatGPT Work and Codex for eligible customers, while the VentureBeat report notes that it is not yet available in regular Chat.
GPT-6 Astra Ultrafast is available through the API, while GPT-6.1 Sol Ultrafast is expected to follow.
Together, the two announcements give developers more flexibility around three key factors in AI deployment: model capability, operating cost and response speed. GPT-6.1 Sol focuses on bringing higher-end performance closer to the cost of a lower-tier model, while Ultrafast gives developers an option to pay more when faster inference is a priority.
