🔍 Read the full analysis: OpenAI Reduces GPT‑6 Sol And Luna Prices By Half—Benchmark Scores Unaffected on ThorstenMeyerAI.com
Get the little things that make your day delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
OpenAI has announced a 50% price reduction for its GPT‑6 Sol and Luna models, effective September 22, 2026. Despite the lower costs, benchmark scores remain largely unchanged, though some quality regressions are noted. This shift significantly impacts AI deployment economics.
OpenAI has reduced the prices of its GPT‑6 Sol and GPT‑6 Luna models by 50%, effective immediately on September 22, 2026, without affecting their benchmark scores. This move aims to democratize access to advanced AI capabilities by lowering operational costs, especially for businesses integrating AI into workflows and products.
OpenAI announced the price cuts as part of its strategy to extend the reach of its latest AI models, GPT‑6 Sol and Luna. The new prices are $2.00 per 1 million tokens for input and $10.00 for output tokens for Sol, and $0.10 per input and $0.50 per output for Luna. These are roughly half the previous costs associated with GPT‑5.6 models, driven by improvements in caching and inference technology, which allow OpenAI to serve these models more cheaply.
Independent analysis from Artificial Analysis confirms that the cost per task has roughly halved, with minimal impact on benchmark scores. Sol at maximum effort scores 48 on the Artificial Analysis Intelligence Index, well above the median of 25, while Luna scores 37 against a median of 12. Despite the cost reductions, some evaluation metrics, particularly in knowledge work, show regressions, attributed to changes in output presentation quality. The models also produce fewer hallucinations, with Sol reducing hallucination rates from 92% to 60%, and Luna from 93% to 77%, partly by declining to answer more often. This trade results in slightly lower accuracy but fewer false answers.
GPT‑6 Sol and Luna: half the price, about the same intelligence
OpenAI’s September 22, 2026 release doesn’t raise the ceiling. It lowers the cost of everything below it, which changes what’s worth automating.
Per 1M input / output tokens. Cached input reads keep the 90% discount.
Cost per task, halved
Measured by Artificial Analysis as the weighted cost of one Intelligence Index task, at max effort.
The effort dial moves cost more than the model choice
| Model and effort | Intelligence Index | Cost per task |
|---|---|---|
| GPT‑6 Sol (max) | 48 | $1.06 |
| GPT‑6 Sol (low) | 34 | $0.13 |
| GPT‑6 Luna (max) | 37 | $0.07 |
| GPT‑6 Luna (low) | 21 | $0.0045 |
| GPT‑6 Luna (non‑reasoning) | 18 | $0.01 |
Sol at low effort keeps about 70% of its max score for roughly an eighth of the cost, because it writes far fewer reasoning tokens. For reference, Claude Opus 5.5 leads the same index at 58.
What got better, and what got worse
Better
- Hallucination rate on AA‑Omniscience: Sol 92% → 60%, Luna 93% → 77%
- Coding Agent Index: Sol 57, up 2 points, at ~50% lower cost per task
- OpenAI reports about half as many factual mistakes for Sol as its predecessor
- Higher cache hit rates; GitHub reports over 50% fewer prompt tokens needing fresh processing
Sol gets there partly by declining more: it attempts 83% of questions vs 99%, and accuracy falls 59% → 54%.
Worse
- GDPval‑AA v2.1: Sol down ~100 Elo, Luna down ~75
- AA‑Briefcase v1.1: Luna down ~45 Elo
- Coding Agent Index: Luna 41, down 2 points
- Both models write more output tokens per task than their predecessors
Reviewers attribute the drops to weaker presentation and deliverables that omit required elements.
What to do about it
Economic Impact on AI Deployment and Cost Efficiency
The price reductions make AI models more accessible for a broader range of applications, especially in sectors where cost constraints previously limited adoption. For businesses, the lower costs could enable more extensive automation, research, and customer support solutions, potentially accelerating AI integration across industries. However, the observed regressions in some knowledge tasks suggest that cost savings may come with trade-offs in output quality for certain workflows, emphasizing the need for testing before deployment.
As an affiliate, we earn on qualifying purchases.
Pricing and Performance Trends in AI Models
OpenAI’s move follows a broader industry trend of reducing model costs to democratize AI access. The release of GPT‑6 Astra two weeks prior set a high-performance benchmark, but the new Sol and Luna models focus on cost efficiency rather than pushing the boundaries of intelligence. The models’ prices are now roughly 50% less than GPT‑5.6, with improvements in caching and inference technology enabling these reductions. Independent evaluations show that while costs are halved, benchmark scores remain stable or improve slightly in some areas, though some knowledge-based tasks have regressed, possibly due to changes in output formatting and presentation.
affordable AI chatbot development tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unresolved Questions About Long-Term Performance
It is not yet clear how these models will perform over extended use cases, especially in complex knowledge tasks or in production environments. The noted regressions in some benchmarks suggest that further testing is needed to understand the full impact of the changes, particularly regarding output quality and reliability in real-world applications. Additionally, the long-term effects of reduced presentation quality on user satisfaction and task accuracy remain to be seen.
GPT-6 compatible AI development kits
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Next Steps for Adoption and Model Evaluation
Businesses and developers are encouraged to test the new models in their specific workflows to assess performance trade-offs. OpenAI may release further updates or refinements based on user feedback and ongoing evaluations. Monitoring how these models perform in large-scale deployments will be crucial, as will any future pricing adjustments or feature enhancements announced by OpenAI.
As an affiliate, we earn on qualifying purchases.
Key Questions
Why did OpenAI reduce the prices of GPT‑6 Sol and Luna?
OpenAI lowered prices primarily through improvements in caching and inference technology, aiming to make AI more affordable and accessible for a wider range of applications.
Do the lower prices affect the models’ performance?
According to independent evaluations, benchmark scores remain stable or improve slightly in some areas, though some knowledge and presentation quality metrics have regressed, indicating possible trade-offs in output presentation.
Are there any risks associated with using these cheaper models?
Potential risks include reduced output quality in certain tasks and an increased likelihood of less detailed or incomplete responses, especially in complex knowledge workflows. Testing is recommended before full deployment.
Will the model prices stay low permanently?
Future pricing depends on ongoing technological improvements, market conditions, and OpenAI’s strategic decisions. No guarantees are given about long-term stability of prices.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
