OpenAI’s Budget-Friendly GPT‑6 Sol And Luna: Benchmark Results Hold Steady
AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: OpenAI’s Budget-Friendly GPT‑6 Sol And Luna: Benchmark Results Hold Steady on ThorstenMeyerAI.com

Prime Big Deal Days · Oct 6–7Offer from Amazon

Get smart everyday buys 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
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

TL;DR

OpenAI has launched GPT‑6 Sol and Luna, two new models priced at half the cost of previous versions. Benchmark results show performance remains stable, with notable reductions in hallucination rates and operational costs. The release aims to make AI more accessible for a range of applications.

OpenAI has introduced GPT‑6 Sol and GPT‑6 Luna, two new language models that are priced at approximately half the cost of their GPT‑5.6 predecessors. The models, launched on September 22, 2026, are designed to expand AI accessibility by offering similar performance at significantly lower prices, marking a shift in how AI capabilities can be integrated into products and workflows.

Both GPT‑6 Sol and Luna are part of OpenAI’s effort to democratize advanced AI by reducing operational costs. GPT‑6 Sol’s input cost per 1 million tokens is now $2.00, down from $4, while Luna’s input cost is $0.10, down from $0.20. Output costs have also been halved, with Sol at $10.00 per 1 million tokens and Luna at $0.50, compared to previous prices of $20 and $1.20 respectively. These reductions are attributed to improvements in caching and inference techniques, which allow the models to operate more efficiently.

Benchmark evaluations by Artificial Analysis, published the same day, confirm that the models maintain comparable performance levels despite the lower costs. Sol scores 48 on the Artificial Analysis Intelligence Index, well above the median of 25 for models in its price class, with Luna scoring 37 against a median of 12. Both models also show improvements in hallucination reduction, with Sol decreasing hallucination rates from 92% to 60%, and Luna from 93% to 77%. However, some performance regressions were noted in knowledge-based tasks, with declines in certain economic and knowledge-work benchmarks, attributed to changes in output presentation quality.

At a glance
updateWhen: announced September 22, 2026
The developmentOpenAI announced the release of GPT‑6 Sol and Luna on September 22, 2026, with benchmark results indicating stable performance despite significant price reductions.

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.

GPT‑6 Sol
$4 / $20 → $2 / $10
GPT‑6 Luna
$0.20 / $1.20 → $0.10 / $0.50

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.

GPT‑5.6 Sol
$1.99
GPT‑6 Sol
$1.06
GPT‑5.6 Luna
$0.18
GPT‑6 Luna
$0.07

The effort dial moves cost more than the model choice

Model and effortIntelligence IndexCost 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

Already on GPT‑5.6 Sol or Luna? The move is mostly a price cut. Re‑test first if your output is a document someone reads, not data a system consumes.
Shelved an automation on cost? Token prices halved and the effort dial adds another order of magnitude. Re‑run the business case.
Choosing between labs? The question is no longer which model is smartest, but which clears your quality bar at the lowest cost per task.
ThorstenMeyerAI.comSources: OpenAI (pricing, vendor benchmarks) and Artificial Analysis (independent evaluation and model pages). Figures as of 23 September 2026.

Impact of Lower-Cost Models on AI Deployment

The release of GPT‑6 Sol and Luna at half the previous prices significantly broadens the range of applications and organizations that can afford to incorporate advanced AI. By maintaining performance levels while reducing costs, OpenAI enables more extensive automation, customer support, research, and content generation tasks. This shift could accelerate AI adoption across industries, especially for smaller firms or projects with tight budgets. However, some performance regressions in specific knowledge tasks highlight the importance of testing models within particular workflows before full deployment.

Amazon

AI language model API access

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Background of OpenAI’s Model Pricing Strategy

OpenAI’s earlier models, including GPT‑5.6, were priced higher, limiting accessibility for smaller organizations. The company’s focus on improving inference efficiency and caching techniques has allowed it to reduce costs without sacrificing overall performance. The launch of Astra, the top-tier model, continues to target high-end applications, but the new Sol and Luna models aim to serve a broader market segment by offering a more affordable entry point into powerful AI capabilities. This approach aligns with OpenAI’s broader goal of democratizing AI technology.

Amazon

affordable AI chatbot development tools

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Unresolved Questions About Model Performance

While benchmark results indicate stable performance, some regressions in knowledge-based tasks suggest that the models may have trade-offs in certain contexts. It remains unclear how well these models will perform in long-term, complex workflows or specialized domains outside the evaluated benchmarks. Additionally, the impact of reduced presentation quality on real-world output consistency needs further observation.

Amazon

AI model cost reduction software

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Next Steps for Adoption and Evaluation

Organizations interested in deploying GPT‑6 Sol and Luna should conduct thorough testing within their specific workflows, especially for knowledge-intensive tasks. OpenAI is expected to release more detailed performance data and user feedback over the coming months. Meanwhile, competitors may respond with their own cost-efficient models, potentially shifting the competitive landscape of AI technology.

Amazon

enterprise AI integration solutions

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Key Questions

How much cheaper are GPT‑6 Sol and Luna compared to previous models?

GPT‑6 Sol’s input cost is $2.00 per million tokens, and Luna’s is $0.10, both approximately half the price of their GPT‑5.6 predecessors. Output costs are similarly halved.

Do the new models perform as well as previous versions?

Benchmark results show that performance remains stable overall, with some improvements in hallucination reduction. However, some knowledge-based tasks experienced regressions, so testing within specific use cases is recommended.

What are the main technical improvements enabling cost reductions?

OpenAI improved caching and inference techniques, which reduce operational costs by enabling more efficient reuse of context and faster processing.

Are there any downsides to these models?

Some evaluations indicate regressions in presentation quality and knowledge-task performance, which could affect workflows requiring detailed, complete outputs.

When can organizations expect more updates or new features?

OpenAI is likely to release further performance data and user feedback in the coming months, with potential updates based on early deployment experiences.

Source: ThorstenMeyerAI.com

This content is for general information only and is not financial, tax or legal advice. Consult a qualified professional for decisions about your money.
EVERGREEN BESTSE

Evergreen bestsellers Picks

As an affiliate, we earn on qualifying purchases.

You May Also Like

How Covert Funding Shaped The 80S Tech Landscape: NeXT And The CIA

Declassified documents reveal CIA covert funding helped sustain NeXT in the 1980s, influencing the early tech landscape and future innovations.

10 Best Home Theater Projectors In 2026

Discover the best home theater projectors of 2026, featuring top models for various needs, including brightness, resolution, and smart features.

How To Accurately Value And Resell Used AI Hardware And GPUs

A manual valuation approach for used GPUs and AI hardware aims to create transparent pricing benchmarks for brokers and resellers.

Revolutionizing Art: 8 Best AI Drawing Tablets In 2026

Discover the 8 best AI-powered drawing tablets in 2026, blending advanced features and affordability for artists of all levels.