--- title: Advancing the price-performance frontier with GPT‑5.6 description: OpenAI is cutting prices for GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20%, effective July 30. Luna now costs $0.20 per million input tokens and $1.20 per... canonical: https://daily.dev/posts/advancing-the-price-performance-frontier-with-gpt-5-6-kn3eddfk3 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Advancing the price-performance frontier with GPT‑5.6 | daily.dev og:description: OpenAI is cutting prices for GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20%, effective July 30. Luna now costs $0.20 per million input tokens and $1.20 per... og:url: https://daily.dev/posts/advancing-the-price-performance-frontier-with-gpt-5-6-kn3eddfk3 og:image: https://api.daily.dev/og/posts/Kn3eDDFk3.png og:image:alt: Advancing the price-performance frontier with GPT‑5.6 og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Advancing the price-performance frontier with GPT‑5.6 **[OpenAI](https://daily.dev/sources/openai)** · 5 min read · 2 upvotes · 1 comments ## Summary OpenAI is cutting prices for GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20%, effective July 30. Luna now costs $0.20 per million input tokens and $1.20 per million output tokens; Terra costs $2 per million input and $12 per million output. A new Fast mode for GPT-5.6 Sol replaces Priority Processing, delivering up to 2.5× faster speeds at twice the standard price. These changes stem from efficiency improvements across model architecture, inference systems, and agentic tooling — including Sol autonomously optimizing production kernels (reducing serving costs 20%) and improving token-generation efficiency by 15%. The GPT-5.6 family is designed to let enterprises mix models across workflow stages, using Sol for complex reasoning and Luna for high-volume, well-specified tasks at a fraction of the cost. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Community take How the wider developer community reacted, aggregated from 1 discussion and 316 comments across hackernews (as of 2026-07-30). **TL;DR:** The community is broadly impressed by the 80% price cut for GPT-5.6 Luna, seeing it as a major competitive move driven by hardware innovation and pressure from Chinese models, though some note Luna was previously overpriced relative to older mini-tier models and debate whether the cuts are truly efficiency-driven or investor-subsidized. **Sentiment:** 55% positive · 30% mixed · 15% skeptical **The case for** - Vera Rubin and other next-gen hardware promise further 10x+ improvements in token throughput per megawatt, suggesting the price floor is still far off. - Cheaper Luna enables running many more parallel agents (e.g., 50 instead of 10) for hypothesis generation and research pipelines at the same cost. - Luna is now competitive with Gemini Flash-tier models at a lower price, making it a strong choice for sub-agent and retrieval workloads. - Third-party inference providers matching DeepSeek's own prices suggests at least some of these cuts reflect real efficiency gains, not pure subsidization. - Burning model weights into silicon (as taalas.com has done for smaller models) could yield orders-of-magnitude further improvements. **The pushback** - Luna had a price hike vs. the older mini model it replaced, so the cut partly just restores prior pricing rather than representing a new floor. - Luna's cached-input price ($0.02/mtok) is still 5x more expensive than DeepSeek V4 Pro's ($0.0036/mtok), limiting its appeal for heavy agentic workloads. - Some argue the price drop may be investor-subsidized market-capture against Anthropic rather than genuine efficiency gains. - Luna struggles with vague prompts and tends to go off-track compared to Sonnet, limiting its use for general-purpose tasks. - Poor-quality cheaper models can litter codebases with wrong comments that later trip up smarter models. **By community** - hackernews (positive): Commenters are broadly excited about the price drop and its implications for multi-agent pipelines, hardware trends, and competitive dynamics, with a vocal minority questioning whether it reflects real efficiency or subsidized pricing. **Hottest debate:** Whether the 80% price cut reflects genuine hardware/software efficiency gains or is investor-subsidized market-capture aimed at undercutting Anthropic. **Open questions** - What is the actual cached-input pricing for Luna after the cut, and how does it compare to DeepSeek for agentic workloads? - Is Luna truly comparable to Sonnet 5 on general tasks, or only on narrow benchmarks? - How quickly will Vera Rubin and next-gen hardware actually roll out and displace existing inference infrastructure? - Will other inference providers (third-party) match OpenAI's new Luna pricing, which would confirm the cuts are cost-driven? **Highlights** > > The kernel work helped reduce the end-to-end cost of serving the model by 20%, while its experiments increased token-generation efficiency by more than 15%. If the cost of serving GPT-5.6 just dropped by 20%, does that add up to literally billions of dollars in savings per month? We know Anthropic spend $1.25 billion renting inference capacity from SpaceX (in two Colossus datacenters) from the SpaceX IPO, but we don't know how much of Anthropic's inference capacity that is (presumably a small fraction, since they were operating on top of AWS and other providers before the SpaceX deal.) I've not seen any numbers that hint at OpenAI's per-month inference bill, but surely that has to be in the multiple billions of dollars as well. So 20% is a really, really big deal. > — [simonw on hackernews · 3 comments](https://news.ycombinator.com/item?id=49113055) > Here is a rough approximation of the pipeline I use: Phase 1 - Run X copies of Luna in parallel over the user's prompt. The purpose is to generate a diverse set of hypotheses. Phase 2 - Run Y copies of Terra in parallel to investigate the hypothesis results, with each receiving them in a randomized order. Phase 3 - Run 1 copy of Sol over investigation reports. The goal is to ensure that the agent covers more initial starting points before presenting a final conclusion. If you only run a single copy of Sol and it hooks onto something wrong, it might not recover. > — [bob1029 on hackernews](https://news.ycombinator.com/item?id=49113955) > > Seeing spikes like this makes me question about where the floor really is. You mean they increased the price and then cut it back and now it is amazing? Luna had a price hike vs mini (its previous replacement). The cut now just puts it back in that ball park. Not that this isn't good news, but what's impressive? > — [re-thc on hackernews · 2 comments](https://news.ycombinator.com/item?id=49113646) > DeepSeek v4 Pro & MiMo v2.5 Pro (Opus 4.6 quality models for code) are insanely cheap for agent-driven work due to their super low cached-input prices ($0.0036/mtok) [0]. For Luna, the cached-input price drop isn't disclosed in TFA, but the pricing page puts it at $0.02/mtok, & that's 5x more expensive. [0] I am constantly surprised how much work pay-as-you-go with DeepSeek / MiMo will get done. I've barely crossed $2 each in a month of use (~200m tokens). > — [ignoramous on hackernews · 1 comments](https://news.ycombinator.com/item?id=49114167) > https://taalas.com/ has done it already for a wildly obsolete model. 14000 tokens per second. https://chatjimmy.ai/ is their interactive. Tiny context, very dumb, but absurdly fast. Imagine this as a tool call for claude code for trivial changes - the tool call from the harness takes longer than the execution. > — [jaggederest on hackernews · 2 comments](https://news.ycombinator.com/item?id=49113868) **Source threads** - [hackernews](https://news.ycombinator.com/item?id=49112867) · 194 points · 316 comments ## Community discussion Top comments from developers on daily.dev. **@petecapecod** · 0 upvotes > cool they're smaller models are really good, so this should be pretty awesome. ## Similar posts on daily.dev - [GPT-5.6 Pricing 2026: Sol, Terra and Luna Tiers Explained](https://daily.dev/posts/gpt-5-6-pricing-2026-sol-terra-and-luna-tiers-explained-teddhg3xw) · finout · 1 upvotes · 0 comments --- Tags: [#ai](https://daily.dev/tags/ai), [#ai-inference](https://daily.dev/tags/ai-inference), [#llm](https://daily.dev/tags/llm), [#openai](https://daily.dev/tags/openai) [View this post on daily.dev](https://daily.dev/posts/advancing-the-price-performance-frontier-with-gpt-5-6-kn3eddfk3) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Advancing the price-performance frontier with GPT‑5.6","url":"https://daily.dev/posts/advancing-the-price-performance-frontier-with-gpt-5-6-kn3eddfk3","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/advancing-the-price-performance-frontier-with-gpt-5-6-kn3eddfk3"},"datePublished":"2026-07-30T17:12:58.356Z","dateModified":"2026-10-14T02:24:52.663Z","description":"OpenAI is cutting prices for GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20%, effective July 30. Luna now costs $0.20 per million input tokens and $1.20 per...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/0b7a2d0b165a2f5acbbf92641515dcc3?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/0b7a2d0b165a2f5acbbf92641515dcc3?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"OpenAI","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"OpenAI","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/984b3632316f4afaa3d5213d5c64fe65","url":"https://daily.dev/sources/openai"},"commentCount":1,"discussionUrl":"https://daily.dev/posts/advancing-the-price-performance-frontier-with-gpt-5-6-kn3eddfk3","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":1}],"keywords":"ai,ai-inference,llm,openai","timeRequired":"PT5M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"OpenAI","item":"https://daily.dev/sources/openai"},{"@type":"ListItem","position":3,"name":"Advancing the price-performance frontier with GPT‑5.6"}]} {"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/advancing-the-price-performance-frontier-with-gpt-5-6-kn3eddfk3","comment":[{"@type":"Comment","text":"cool they’re smaller models are really good, so this should be pretty awesome.","datePublished":"2026-07-31T12:56:40.137Z","url":"https://daily.dev/posts/Kn3eDDFk3#c-3B1g4fAis","author":{"@type":"Person","name":"Peter Cruckshank","url":"https://daily.dev/petecapecod","image":"https://media.daily.dev/image/upload/s--ZJhQyKws--/f_auto/v1721235024/avatars/avatar_A9xh33q0QoxtkGoJRCosp"}}]} ```