OpenAI Launches GPT-6.1 Sol: Near-Astra Intelligence at One-Fifth the Compute Cost.

OpenAI Releases GPT-6.1 Sol: Near-Astra Level Intelligence at One-Fifth the API Cost

OpenAI has officially deployed its latest mid-tier model release, GPT-6.1 Sol, upgrading its popular flagship-adjacent architecture with performance metrics that closely approach its top-tier GPT-6 Astra model at just one-fifth (1/5) of Astra's operational cost. The updated model bridges the gap between high-end reasoning capabilities and cost-effective enterprise execution across coding, complex document processing, and automated agent workflows.

Benchmark Metrics, Pricing Architecture, and Availability

The GPT-6.1 Sol release delivers significant gains in token efficiency, caching discounts, and specialized workflow execution across enterprise environments:

• Benchmark Performance Highlights:

  • DeepSWE v1.1 Coding Benchmark: Scores nearly identically to OpenAI’s premier GPT-6 Astra, demonstrating high-level software engineering and autonomous debugging capabilities.

  • GDP.pdf Document Analysis: Outperforms Anthropic’s Claude Opus 5.5 in processing, comprehending, and synthesizing complex multi-page document structures.

  • AutomationBench Workflows: Slightly surpasses Claude Opus 5.5 in multi-step task execution, system orchestration, and agentic process automation while running at a fraction of the compute expense.

• API Cost Architecture & Caching Discounts:

  • Standard Rates: Maintained at the accessible GPT-6 baseline rate of $2.00 per million input tokens and $10.00 per million output tokens.

  • Prompt Caching Discount: Cached input pricing has been cut in half to $0.10 per million tokens (a 50% reduction compared to the previous GPT-6 Sol generation).

• Deployment Rollout & API Access:

  • Supported Platforms: Rolling out immediately across ChatGPT Work and Codex for paid subscription tiers, including Plus, Pro, Business, Enterprise, and Edu users (standard consumer web chat access is not enabled at launch).

  • Developer API Endpoint: Available via the gpt-6.1-sol API endpoint.

  • Upcoming Variant: OpenAI confirmed that a specialized high-speed variant, GPT-6.1 Sol Ultrafast, is in development, offering significantly reduced latency while supporting an 8x expanded token throughput capacity.

 

Source: OpenAI 

💬 AI Content Assistant

Ask me anything about this article. No data is stored for your question.

OpenAI Releases GPT-6.1 Sol: Near-Astra Level Intelligence at One-Fifth the API Cost

OpenAI has officially deployed its latest mid-tier model release, GPT-6.1 Sol, upgrading its popular flagship-adjacent architecture with performance metrics that closely approach its top-tier GPT-6 Astra model at just one-fifth (1/5) of Astra's operational cost. The updated model bridges the gap between high-end reasoning capabilities and cost-effective enterprise execution across coding, complex document processing, and automated agent workflows.

Benchmark Metrics, Pricing Architecture, and Availability

The GPT-6.1 Sol release delivers significant gains in token efficiency, caching discounts, and specialized workflow execution across enterprise environments:

• Benchmark Performance Highlights:

  • DeepSWE v1.1 Coding Benchmark: Scores nearly identically to OpenAI’s premier GPT-6 Astra, demonstrating high-level software engineering and autonomous debugging capabilities.

  • GDP.pdf Document Analysis: Outperforms Anthropic’s Claude Opus 5.5 in processing, comprehending, and synthesizing complex multi-page document structures.

  • AutomationBench Workflows: Slightly surpasses Claude Opus 5.5 in multi-step task execution, system orchestration, and agentic process automation while running at a fraction of the compute expense.

• API Cost Architecture & Caching Discounts:

  • Standard Rates: Maintained at the accessible GPT-6 baseline rate of $2.00 per million input tokens and $10.00 per million output tokens.

  • Prompt Caching Discount: Cached input pricing has been cut in half to $0.10 per million tokens (a 50% reduction compared to the previous GPT-6 Sol generation).

• Deployment Rollout & API Access:

  • Supported Platforms: Rolling out immediately across ChatGPT Work and Codex for paid subscription tiers, including Plus, Pro, Business, Enterprise, and Edu users (standard consumer web chat access is not enabled at launch).

  • Developer API Endpoint: Available via the gpt-6.1-sol API endpoint.

  • Upcoming Variant: OpenAI confirmed that a specialized high-speed variant, GPT-6.1 Sol Ultrafast, is in development, offering significantly reduced latency while supporting an 8x expanded token throughput capacity.

 

Source: OpenAI