OpenAI launches two more GPT-6 models: Sol and Luna, with prices reduced by 50%, accelerating the deployment of Astra capabilities.

The GPT-6 model family is expanding rapidly.

On Tuesday, September 22nd, Eastern Time, OpenAI announced the release of two new models, GPT-6 Sol and GPT-6 Luna. Less than three weeks after OpenAI released its flagship model, GPT-6 Astra, which it claims is the "world's most intelligent," OpenAI has further expanded its GPT-6 product line and brought some of Astra's capabilities to faster, cheaper models.

Another key highlight of these two models is their price: OpenAI announced that the API prices of GPT-6 Sol and GPT-6 Luna are 50% lower than the promotional price of GPT-5.6. OpenAI stated that this is due to further improvements in caching and inference efficiency, directly passing the cost savings on to users.

According to OpenAI's announcement, Sol and Luna are now available to Plus, Pro, Business, Enterprise, and Edu users in ChatGPT Work and Codex, with their APIs also open. Free and Go users can experience GPT-6 Luna in the desktop application. However, neither model is currently available in the regular ChatGPT chat interface.

Less than three weeks after Astra's release, OpenAI adds two more GPT-6 compliant technologies.

Judging from the release schedule of GPT-6, OpenAI's actions this time are quite frequent.

On September 3, OpenAI officially launched GPT-6 Astra, claiming it represents a new frontier in capabilities across areas such as computer and browser operation, software engineering, cybersecurity, and scientific and professional work. OpenAI positioned Astra as the flagship model of the GPT-6 family at the time.

In less than three weeks, OpenAI launched Sol and Luna.

OpenAI’s official social media account stated that Sol and Luna “build on the technological advancements behind GPT-6 Astra”, bringing Astra’s many advantages to “faster, more affordable models” to support large-scale work.

This product rollout shows that OpenAI did not position GPT-6 as a single upgrade to a flagship model, but rather rapidly expanded its product line to different performance, cost, and use cases after the release of Astra.

This also echoes the increasingly clear product strategy of AI model manufacturers: flagship models are responsible for pushing the limits of capabilities, while faster and cheaper versions are responsible for entering more high-frequency workflows.

From Astra to Sol and Luna, capabilities are being extended to larger-scale scenarios.

Unlike Astra, which emphasizes cutting-edge capabilities, Sol and Luna are positioned to prioritize speed, cost, and scalability .

OpenAI states that Sol and Luna not only build upon Astra's technological advancements but also optimize caching and inference efficiency. The company claims these optimizations reduce model runtime costs and ultimately translate into lower API prices for users.

OpenAI also stated that both models have incorporated Astra's advancements in alignment capabilities, showing improvements over their respective GPT-5.6 versions.

In other words, OpenAI is not simply launching two "low-cost" models, but is attempting to extend some of Astra's advancements in capabilities and alignment to more frequent and larger-scale model invocation scenarios.

For developers and businesses, this means that advanced model capabilities, which were previously only worth using in high-value tasks, may be able to be applied to a large number of repetitive tasks such as coding, intelligent agents, and enterprise automation at a lower unit cost in the future.

API prices have been halved compared to the promotional price of GPT-5.6.

Price is the clearest business signal from this release.

OpenAI announced that the API prices for GPT-6 Sol and GPT-6 Luna have been reduced by 50% compared to the promotional price of GPT-5.6 .

OpenAI's current promotional price for GPT-5.6 Sol is $4 per 1 million tokens input and $20 per 1 million tokens output; therefore, this price reduction for Sol and Luna signifies a further decrease in their API costs. OpenAI's corporate pricing list also shows that GPT-5.6 Sol is currently still in the promotional pricing phase.

OpenAI explained that the price reduction mainly came from more efficient caching and inference, rather than simply a business concession.

OpenAI's official account stated that "higher usage limits and lower costs" will give users more flexibility and room for iteration.

For developers and enterprises that need to call models extensively, the decrease in unit token cost means that more tasks can be run with the same budget, and it also means that intelligent agent workflows that were previously limited by cost are expected to gain greater operational space.

This has extended the competition in AI models from "whose model is more powerful" to "who can provide a sufficiently powerful model at a lower cost" .

First, go to Work and Codex; free users can try Luna.

In terms of product coverage, OpenAI did not immediately integrate Sol and Luna into the ChatGPT regular chat interface.

According to OpenAI's arrangements, Plus, Pro, Business, Enterprise, and Edu users can use two models in ChatGPT Work and Codex, and also call them via API.

Free and Go users can experience GPT-6 Luna in the desktop application.

This arrangement makes the initial launch scenarios for both models clearly biased towards work and development , rather than simply consumer-facing chat.

Among them, Sol and Luna's low cost and high efficiency make them more suitable for coding tasks that require frequent model calls, agent workflows, and enterprise internal automation.

In contrast, GPT-6 Astra still serves as the flagship model. After Astra's release, OpenAI even temporarily suspended new Pro subscription registrations due to the surge in demand, stating that Astra's demand was "unprecedented."

Therefore, the launch of Sol and Luna may help OpenAI redirect some high-frequency demands to lower-cost models, thereby reducing infrastructure pressure while expanding usage.

The competition in AI models has shifted from "competing for flagship models" to "competing for scale."

The release of Sol and Luna also occurred against the backdrop of AI model vendors launching cheaper versions in quick succession.

Also on Tuesday, OpenAI's rival Anthropic launched its new model, Claude Opus 5.5, emphasizing lower operating costs. The two AI giants released new, more cost-competitive models almost simultaneously. The focus of market competition is shifting from simply pursuing the upper limit of model capabilities to balancing performance, speed, and price.

For OpenAI, GPT-6 Astra is responsible for setting the upper limit of capabilities, while Sol and Luna are tasked with pushing these technological advancements to a wider range of work scenarios.

From the release of Astra on September 3, to the launch of Sol and Luna less than three weeks later, the GPT-6 family is rapidly shifting from "flagship model releases" to "product line expansion".

The further 50% price reduction of the API compared to the promotional price of GPT-5.6 sends a clear signal: as model capabilities gradually become more homogenized, OpenAI not only wants to compete for the "number one" title of the strongest model, but also for the large-scale market of how many times developers and enterprises are truly willing to call the model every day.

Risk Warning and DisclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.