OpenAI Launches GPT-6.1 Sol, Offering Near-Flagship AI Performance at Sharply Lower Cost

OpenAI announced on Tuesday the release of GPT-6.1 Sol, a new artificial intelligence model designed to offer performance comparable to its flagship GPT-6 Astra at a significantly reduced price point. The launch aims to make high-end AI capabilities more accessible to developers and enterprise users, potentially shifting the competitive landscape for AI services.

A New Cost-Performance Benchmark

According to OpenAI, the primary advantage of GPT-6.1 Sol is its cost-efficiency. The model is priced at one-fifth of the cost of the more powerful GPT-6 Astra, creating a new mid-tier option for customers balancing performance needs with budget constraints.

The company detailed the new pricing structure:

• Input Tokens: $2 per million
• Output Tokens: $10 per million
• Cached Input Tokens: $0.10 per million

This compares to Astra's rates of $10 and $50 for input and output tokens, respectively. The cached input rate for GPT-6.1 Sol also represents a 50% reduction from the original GPT-6 Sol model, further lowering operational costs for frequent, repetitive tasks.

Performance Across Key Evaluations

Ad

OpenAI released several benchmark results to demonstrate the new model's capabilities. On the DeepSWE v1.1 software engineering benchmark, GPT-6.1 Sol reportedly matched the performance of GPT-6 Astra while maintaining its lower cost structure. It also scored 6.4 percentage points higher than its predecessor, GPT-6 Sol, on the same evaluation.

In tests involving complex document analysis (GDP.pdf), GPT-6.1 Sol outperformed the competing Opus 5.5 model at less than half the cost per task. For computer use tasks measured by OSWorld 2.0, the new model came within 2.1 percentage points of Astra’s score at just one-seventh of the cost. OpenAI also noted a 32% reduction in factual errors on deliberately difficult prompts compared to GPT-6 Sol, with the error rate falling from 11.4% to 7.7%.

Availability and Market Impact

The new model is immediately available to Plus, Pro, Business, Enterprise, and Edu users through ChatGPT Work and Codex. Developers can access it via the OpenAI API under the name `gpt-6.1-sol`. By providing near-top-tier performance at a substantially lower price, OpenAI is positioning GPT-6.1 Sol to capture a wider segment of the market and challenge competitors on both capability and price.

Looking ahead, OpenAI stated it will release GPT-6.1 Sol Ultrafast in the coming days, a variant promising up to eight times faster token generation. This move suggests a strategy of segmenting its offerings to address diverse user needs, from high-accuracy analysis to high-speed, real-time applications.