OpenAI launches GPT-6 Sol and Luna models with 50% lower API costs FAST-MOVING

OpenAI GPT-6 Sol GPT-6 Luna AI API pricing artificial intelligence

OpenAI has expanded its GPT-6 model family with the release of GPT-6 Sol and GPT-6 Luna, designed as more efficient and affordable alternatives to the flagship GPT-6 Astra. These new models are specifically optimized for professional tasks, including coding, automation, and computer-use applications.

Both models feature API pricing that is 50% lower than the promotional rates of their GPT-5.6 predecessors. GPT-6 Sol is priced at $2 per million input tokens and $10 per million output tokens, while GPT-6 Luna is set at $0.10 per million input tokens and $0.50 per million output tokens. OpenAI attributes these cost reductions to improvements in inference infrastructure and prompt caching, the latter of which can reduce costs for cached input tokens by up to 90%.

In terms of performance, OpenAI claims GPT-6 Sol demonstrates significant improvements in factuality, with errors reduced by approximately half compared to GPT-5.6 Sol. The models are being rolled out to ChatGPT Plus, Pro, Business, Enterprise, and Edu users via ChatGPT Work and Codex, with Luna also becoming available to Free and Go users on the desktop application. The release coincides with intense competition in the AI sector, following similar efficiency-focused updates from Anthropic.

Entities

Amazon · Anthropic · GPT-6 Astra · GPT-6 Luna · GPT-6 Sol · Microsoft · OpenAI · Perplexity AI · Sam Altman

Claims

What the coverage asserts, and how many sources carry each claim.

[● 31 SOURCES] OpenAI released two new models, GPT-6 Sol and GPT-6 Luna, as part of the GPT-6 family. www.iphoneincanada.ca · dijitaliyidir.com · unwire.pro · donanimgunlugu.com · www.blogspan.net · +24 more
[● 31 SOURCES] The API pricing for GPT-6 Sol and GPT-6 Luna is 50% lower than their GPT-5.6 predecessors. www.iphoneincanada.ca · dijitaliyidir.com · unwire.pro · donanimgunlugu.com · www.blogspan.net · +24 more
[● 3 SOURCES] GPT-6 Sol's coding deception rate decreased from 10.4% in GPT-5.6 to 1.3%. decrypt.co · www.itmedia.co.jp
[● 27 SOURCES] GPT-6 Luna is priced at $0.10 per million input tokens and $0.50 per million output tokens. www.iphoneincanada.ca · unwire.pro · www.blogspan.net · news.az · the-decoder.com · +20 more
[● 3 SOURCES] Perplexity AI has set GPT-6 Sol as the default model for its 'Light' preset in Effort Mode. unwire.pro · ascii.jp
[● 26 SOURCES] GPT-6 Sol is priced at $2 per million input tokens and $10 per million output tokens. www.iphoneincanada.ca · unwire.pro · www.blogspan.net · news.az · the-decoder.com · +19 more
[● 4 SOURCES] Prompt caching upgrades can slash costs on cached input tokens by up to 90%. www.iphoneincanada.ca · unwire.pro · www.itmedia.co.jp
[● 11 SOURCES] The new models specifically target professional work, coding, automation, and computer-use tasks. bitnewsbot.com · news.az · ascii.jp · dijitaliyidir.com · dominicanaaldia.com · +6 more
[● 2 SOURCES] Anthropic released the Claude Opus 5.5 model, which is designed for complex tasks, software debugging, and financial analysis. geekyalgeria.com · www.prodiris.fr
[○ 1 SOURCE] OpenAI's GPT-6 Astra outperformed human-assisted AI operators in five drone-based surveillance tasks, including 3D environment reconstruction and target tracking. legrandcontinent.eu
[○ 1 SOURCE] In scientific research testing via Terminal-Bench-Science 0.1, Opus 5.5 scored 58.7%, significantly higher than the 29% scored by Opus 5. www.prodiris.fr
[● 2 SOURCES] The Claude Opus 5.5 model costs $4 per million input tokens and $20 per million output tokens, representing a 40% reduction in cost for typical workloads compared to Opus 5. www.prodiris.fr · geekyalgeria.com