🔍 Read the full analysis: What Does It Mean When GPT‑6 Sol And Luna Prices Drop 50% But Benchmarks Stay Flat? on ThorstenMeyerAI.com
Get the latest gadgets delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
TL;DR
OpenAI has reduced the prices of GPT-6 Sol and Luna models by 50%, yet their benchmark scores remain stable. This shift aims to make AI more accessible without sacrificing performance, but some quality regressions are noted.
OpenAI has reduced the prices of its GPT‑6 Sol and Luna models by approximately half, effective September 22, 2026, while their benchmark scores remain largely stable. This move aims to democratize access to AI capabilities by lowering costs for mid-tier models, making automation more viable across industries. The price reductions come amid broader industry efforts to improve AI affordability without compromising quality, but the impact on model performance and reliability remains under scrutiny.
On September 22, 2026, OpenAI announced that its GPT‑6 Sol and Luna models are now priced at 50% less than their GPT‑5.6 predecessors. The new pricing sets GPT‑6 Sol at $2.00 per 1 million input tokens and $10.00 per 1 million output tokens, down from $4 and $20 respectively. Similarly, GPT‑6 Luna now costs $0.10 per 1 million input tokens and $0.50 per 1 million output tokens, halving previous prices.
OpenAI attributes the cost reductions to improvements in caching and inference technologies, which lower operational expenses. The company emphasizes that these models are designed to distribute Astra’s advanced intelligence more broadly, making AI deployment more cost-effective for businesses and developers. Despite the lower prices, performance benchmarks—such as the Artificial Analysis Intelligence Index—show that the models’ capabilities remain steady, with Sol scoring 48 and Luna 37, well above median levels for their class.
Independent evaluations by Artificial Analysis confirm that the cost per task has roughly halved, with no significant decline in overall scores. For example, Sol’s cost per task at maximum effort dropped from about $1.99 to $1.06, and Luna’s from roughly $0.20 to $0.07, despite a slight increase in output tokens per task. Notably, some quality metrics, such as hallucination rates, improved—Sol’s hallucination rate decreased from 92% to 60%, and Luna’s from 93% to 77%. However, there are reports of regressions in knowledge-based tasks, where models scored lower on certain economic and knowledge work benchmarks, potentially due to changes in output presentation quality.
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
Implications for AI Deployment and Cost Management
The price cuts for GPT‑6 Sol and Luna models significantly lower the financial barrier for integrating advanced AI into products and workflows. This shift could accelerate AI adoption across sectors such as customer service, research, and automation, enabling smaller firms to access high-quality language models at a fraction of previous costs. However, some users should be aware of potential trade-offs, including minor regressions in knowledge accuracy and output presentation quality, which may impact applications requiring precise, detailed deliverables. Overall, this move underscores a strategic focus on making AI more accessible and cost-efficient without sacrificing core performance benchmarks, potentially reshaping competitive dynamics in the AI industry.
As an affiliate, we earn on qualifying purchases.
Background on Pricing and Performance Trends
OpenAI’s earlier models, including GPT‑5.6, were priced higher, reflecting their advanced capabilities and larger scale. The release of Astra-based models marked a shift towards emphasizing cost efficiency alongside performance. Prior to this, AI providers generally increased model capabilities with corresponding price hikes, but recent industry trends, including OpenAI’s, show a move toward balancing performance with affordability. The announcement of GPT‑6 Sol and Luna at half-price follows similar industry efforts, such as Anthropic’s reduction in Claude Opus 5.5 prices by 20%. The models’ evaluation scores and benchmarks have remained stable overall, although some specific knowledge tasks have shown slight regressions, highlighting ongoing challenges in balancing cost, quality, and reliability.
Unconfirmed Effects on Long-term Model Reliability
It is not yet clear how sustained use of these lower-cost models will impact long-term reliability and accuracy, especially in high-stakes applications. While initial evaluations show stable benchmarks, some knowledge-based tasks have experienced regressions, and the effects of reduced presentation quality are still being studied. Industry experts are watching for potential shifts in hallucination rates, factual accuracy, and model robustness over extended deployment periods.
Next Steps for Market Adoption and Performance Monitoring
OpenAI is expected to continue monitoring the performance of GPT‑6 Sol and Luna in real-world applications, collecting user feedback and conducting further evaluations. The company may also release updates or fine-tuning options to address any emerging issues. Industry analysts anticipate increased adoption of these models across various sectors due to their lower costs, with some organizations conducting pilot programs to assess long-term reliability and output quality. Additional benchmark releases and independent reviews are likely to follow, providing clearer insights into the models’ evolving capabilities.
Key Questions
Why did OpenAI reduce the prices of GPT‑6 Sol and Luna models?
OpenAI reduced prices due to improvements in caching and inference technologies that lowered operational costs, enabling pass-through savings to customers while maintaining performance.
Do the lower prices mean the models are less capable?
No, independent evaluations show that benchmark scores remain stable, although some specific knowledge tasks have experienced minor regressions. Overall performance levels are comparable to previous models.
Are there any risks associated with using these cheaper models?
Potential risks include slight regressions in knowledge accuracy and output quality, particularly in detailed or complex tasks. Long-term reliability in critical applications remains under observation.
How might this price reduction affect the AI industry?
This move could accelerate AI adoption among smaller firms and new markets, increasing competition and pushing other providers to lower their prices or improve offerings.
Source: ThorstenMeyerAI.com
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
