OpenAI releases GPT-6 Sol and GPT-6 Luna, cuts API prices 50% from GPT-5.6

OpenAI shipped two new models today, GPT-6 Sol and GPT-6 Luna, extending the GPT-6 family down into cheaper, more efficient tiers while cutting API prices in half compared with their GPT-5.6 predecessors.

What's new

GPT-6 Sol and GPT-6 Luna are positioned as the workhorse and budget tiers of the GPT-6 lineup, respectively, sitting below the flagship GPT-6 Astra. Per OpenAI's announcement, "We're expanding the GPT‑6 universe with GPT‑6 Sol and GPT‑6 Luna."

Pricing:

• GPT-6 Sol: $2 per million input tokens, $10 per million output tokens
• GPT-6 Luna: $0.10 per million input tokens, $0.50 per million output tokens

OpenAI says the cuts come from infrastructure gains rather than a promotional discount: "reducing API prices for Sol and Luna by 50% compared with their GPT‑5.6 promotional pricing," adding that "Improvements in caching and inference let us serve these models at lower cost, and we're passing those savings directly on to users." Cached token reads carry a further 90% discount over fresh input pricing.

According to OpenAI, quality moved forward alongside the price cut: "both Sol and Luna show improvements over their GPT‑5.6 counterparts," with the company citing gains in alignment, factuality, and a more concise communication style.

GPT-6 Sol's published spec sheet lists a 1,050,000-token context window (922,000 max input, 128,000 max output tokens), text and image input with text output, an Apr 20, 2026 knowledge cutoff, and support for reasoning effort levels from none through max, plus streaming, structured outputs, function calling, file search, web search, and prompt caching across both the Chat Completions and Responses APIs.

Both models are rolling out today: Sol and Luna are live now in ChatGPT Work and Codex for Plus, Pro, Business, Enterprise, and Edu users, with Luna also reaching Free and Go users through the desktop app. API access under the gpt-6-sol and gpt-6-luna model IDs is rolling out gradually over the course of the day.

Context

Sol and Luna have been OpenAI's naming pattern for its mid-tier and budget-tier models since the GPT-5.6 generation, which itself got a hallucination-rate-focused refresh in August. This release moves the same two-tier structure onto the GPT-6 base while resetting pricing rather than running it as a limited-time promotion — the 50% cut is explicitly framed against GPT-5.6's own promotional rate, meaning list pricing effectively halves going forward.

Why it matters

For developers building on OpenAI's API, this is a direct cost reduction on the two model tiers most likely to run high request volumes — coding agents, chat support, and other agentic workloads where GPT-5.6 Sol and Luna were already the default choice for cost-sensitive traffic. A permanent 50% price cut, stacked with a 90% cached-read discount, meaningfully changes the unit economics for anyone running these models at scale, and keeps pressure on competing labs' mid-tier pricing as the GPT-6 generation rolls out.