GPT-6 Luna
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 17
- PhAI Labs releases DiscoveryRSI research agent and 27B model
PhAI Labs, working with several universities and research institutions, proposed DiscoveryRSI and trained DiscoveryRSI-27B using a hypothesis-credit attribution mechanism, releasing the paper, code and models. The system stores evolving hypotheses and test plans in a research hypothesis tree and is paired with DFM-Bench, which evaluates hypothesis revisions, research paths and whole research episodes. The team reports that DiscoveryRSI's group scored 6.3% in a literature cross-check, the highest of five model groups, while the Claude Opus 5 group scored 2.3% and ranked last; it cautions that group sample sizes differ and the comparison is descriptive, so it does not prove the hypotheses were adopted or validated by later research.
机器之心 · 🔥 9 - Jev becomes Vercel's fastest-adopted AI model ever
Vercel data published on Thursday shows that TypeSafe AI's lightweight decision model Jev was used by 35% of paid teams within two weeks of its limited release, nearly double the reach of the second most popular model, Sonnet 5.5, making it the fastest-adopted model in its first two weeks on Vercel's platform. Released in limited form on September 15, Jev restricts outputs to a few predetermined options to run cheaper and faster, making it suited to classifying large volumes of data; Vercel has used it to replace GPT-6 Luna for classifying harmful commands. The figures come from Vercel's AI gateway, a unified API service similar to OpenRouter, and may undercount labs such as OpenAI and Anthropic that offer their own APIs; the same report says open-source models account for more than 70% of tokens processed through the gateway and nearly 25% of total spend.
TechCrunch AI · 🔥 7 - StarSkirmish Bench v0.2: GPT-6 Astra tops LLM StarCraft bot ranking
StarSkirmish Bench v0.2 gives each large model one hour to write a Protoss bot in C++ for StarCraft: Brood War, then scores it by win rate against human-written and demo bots. GPT-6 Astra leads with 57, followed by GPT-6.1 Sol at 46, with Gemini 4 Argon and Claude Opus 5.5 tied at 39, Claude Fable 5.1 at 37, then Claude Sonnet 5.5 (26), Muse Wonder Full and Claude Haiku 5.5 (21) and GPT-6 Luna (13). Even the best model, GPT-6 Astra, wins only about 7% of games against top human-written bots Stardust, PurpleWave and BananaBrain, so the benchmark is far from saturated; practicing against strong opponents is the strongest predictor of a model's score, with Claude Fable 5.1 a clear outlier.
Hacker News · AI（100+ 分） · 🔥 0 - Tessl ships Verifiers GA and reports 75% of its PRs merge without human review
Tessl announced that its Verifiers feature is generally available: plain-text repository rules are checked on every code change by an LLM judge, returning pass or fail and runnable locally, in pre-commit, or as a CI gate. The default judge is openai/gpt-6-luna, with Jev offered as an alternative. In a separate post, Tessl says about 75% of PRs at its team, which opens over 1,200 agent-written PRs weekly, merged without human correction, and that 91% of Tessl Code Review findings were real defects versus 68% for its previous AI reviewer, at roughly 35% lower review cost per PR; the figures are vendor-reported.
Tessl Blog · 🔥 0 - OpenAI retires GPT-5.5 from ChatGPT as legacy API versions get a three-month sunset
OpenAI removed GPT-5.5 from ChatGPT, ChatGPT Work and Codex sessions signed in with a ChatGPT account on October 14, according to reports citing the company's model documentation; API access was reported as unaffected, and Codex sessions using API-key authentication were said to keep the model. Reports add that selected API versions of GPT-4o, GPT-4o mini, GPT-4.1 and GPT-4.1 mini are scheduled to shut down on January 14, 2027, leaving developers roughly three months to migrate. OpenAI advised users to check workspace default models, custom agents, scheduled tasks and scripts that still specify GPT-5.5, pointing Plus and Pro Codex users to GPT-6.1 Sol and Free and Go users to GPT-6 Luna in supported desktop apps. A 36Kr consumer piece also described Reddit users writing prompt-based "reincarnation" guides to preserve GPT-5.5's tone and conversational style before it left the product.
36氪 人工智能 · 🔥 0 - OpenAI details Third Moment Research's 10,000-agent runs lasting 36-48 hours
OpenAI published a customer case study on how startup Third Moment Research runs a self-improving agent system on its models: 10,000 agents at once inside Docker environments, with individual investigations lasting 36 to 48 hours. The team says Codex working with GPT-5.6 analyzed thousands of recorded simulation trajectories and found that backtesting a year of order-level US equity data took hours; Codex then identified caching and memory-packing fixes and shipped them to production, cutting a backtest from roughly 3 hours to roughly 3 to 5 minutes. Third Moment wraps the inner Codex harness in an outer model driven by GPT-6 Sol or GPT-6.1 Sol that decides whether to retry, terminate or escalate to a human, and uses the Daybreak security platform for sandbox penetration testing; per 36Kr, the setup consumes about 1 trillion tokens a day, with GPT-6.1 Sol the most-used model for its token efficiency.
OpenAI · 🔥 0 - GPT-6 Sol, Luna, Astra on Amazon Bedrock via global cross-Region inference
An AWS Machine Learning Blog post describes how Amazon Bedrock serves OpenAI's GPT-6 Sol, GPT-6 Luna and GPT-6 Astra through global cross-Region inference from Thailand, Malaysia, Singapore, Indonesia and Taiwan, with GPT-6 Astra callable from Asia Pacific (Singapore) as the source Region. It walks through invocation via the Converse API, the OpenAI-compatible Responses and Chat Completions paths, and the ChatGPT desktop app and Codex CLI (version 0.148 or later), along with IAM and service control policy setup plus quota and monitoring notes. The post states that Sol and Luna accept reasoning effort levels from none to max, while Astra rejects none.
AWS Machine Learning Blog · 🔥 0 - Inside the OpenAI–Anthropic feud: Amodei's exit and the rivalry that followed
A retrospective by Chinese outlet Gelonghui (republished by 36Kr) recounts the December 2020 OpenAI all-hands where the screen went black and Dario Amodei left with six core members, an episode internally dubbed "The Divorce," tracing the rift to the 2017 "Countries" AGI-consortium idea, an unwritten promise tied to Microsoft's $1B investment, and OpenAI's refusal to give Amodei's safety team veto power over model releases. It says Anthropic was founded in 2021 with plans for only 50 staff and cites 2026 rivalry figures: a $65B round in May at a $965B post-money valuation versus OpenAI's $852B, Anthropic at $65B annualized revenue by late July against OpenAI's just over $40B, and Menlo Ventures data showing Anthropic ahead on enterprise LLM spend 40% to 27%, with Claude Code at roughly 54% of the coding segment. It also says that on Oct 7 OpenAI pushed GPT-6 to all ChatGPT users (paid tier GPT-6 Sol, free tier GPT-6 Luna) while Anthropic released Claude Haiku 5.5 at $0.10 per million input tokens, about 75% cheaper than the prior generation.
36氪 人工智能 · 🔥 0 - Step 5 Preview free for a week on OpenCode, Cline, Kilo and Nous Portal
StepFun said its flagship model Step 5 Preview is free for one week in OpenCode, Cline, Kilo and Nous Research's Nous Portal for agentic and professional work. Material relayed by StepFun describes the model as a 600B-total, 27B-active MoE with a 1M-token context window and vision, and the OpenCode post also lists zero data retention. Nous Research is quoted as saying Step 5 Preview scored 33.89 on the Hermes Index, matching GPT-6 Luna, while Cline says it scores ahead of Kimi K3 and GLM-5.3 on DeepSWE; these benchmark comparisons are partner or vendor claims that have not been independently verified.
阶跃星辰 StepFun · 🔥 0 - GPT-6 Luna leads Decisions model speed ranking
OpenRouter said on X that GPT-6 Luna is currently the fastest Decisions model, at 180ms on global requests. The post ranks Jev and Perplexity Decider behind it. No measurement methodology or latency figures for the runners-up were provided.
OpenRouter (X) · 🔥 0 - LegalOn halves Codex costs by tiering GPT-6 Astra, Sol and Luna
After embedding Codex in its development process, LegalOn Technologies initially gave engineers unlimited access to GPT-5.5 in Fast mode, which pushed up spending. Its AI-powered Development CoE then introduced model-selection guidelines: GPT-6 Luna for everyday implementation, GPT-6.1 Sol for standard design, data analysis and documents, and GPT-6 Astra for complex analysis, architecture and agent orchestration, alongside restricting Fast mode by default and setting monthly budget caps per department, group and individual. OpenAI says the company cut estimated daily costs by roughly 65% versus GPT-5.5 while maintaining development speed, with mature businesses asked for about 20% better cost efficiency and new businesses given generous budgets.
OpenAI · 🔥 0 - OpenAI opens Decisions API public beta as GPT-6 Luna Decisions lands on OpenRouter
OpenAI's Decisions API is now in public beta for all developers, letting an app pick the right model, tool or action in near real-time by sending text, JSON or images and receiving typed answers with probabilities. OpenRouter has simultaneously listed GPT-6 Luna Decisions at $0.10 per million input tokens with free output and a 1M-token context window. OpenAI says the API makes decisions up to 10x faster than GPT-6 Luna through the Responses API.
OpenRouter (X) · 🔥 0 - Anthropic cuts Claude Sonnet 5.5 cache read price to $0.10, adds API credits
Anthropic has lowered the Claude Sonnet 5.5 cache read price from $0.20 to $0.10 per million tokens and is adding monthly API credits: $100 for Max 5x, $200 for Max 20x, and up to $500 shared among Team users, per VentureBeat. The report says the model starts at $0.10 per million input tokens and $0.50 per million output tokens, matching OpenAI's GPT-6 Luna.
Techmeme · 🔥 0 - OpenAI launches Decisions API for fast choices, targeting Jev
At DevDay, OpenAI announced a Decisions API that narrows open-ended generation to picking from predefined options and returns a choice plus confidence that code can consume directly, which The Decoder says runs about ten times faster than the Responses API. It currently supports only GPT-6 Luna, at $0.10 per million input tokens with free output tokens in public beta, according to The Decoder, while InfoQ describes it as a limited preview with pricing and customization details to come at general availability. OpenAI also cut its paid API tiers from five to three: Build, Launch and Grow.
The Decoder · 🔥 0 - Anthropic launches Claude Haiku 5.5 at $0.10 per million input tokens
Anthropic released Claude Haiku 5.5 on October 7, describing it as the fastest, cheapest and most capable model in its Haiku small-model line. For requests up to 100,000 tokens, it costs $0.10 per million input tokens and $0.50 per million output tokens, rising to $0.50 and $2.50 above that threshold; Anthropic says it costs about 75% less to run on average than Haiku 4.5. Anthropic-reported figures show OSWorld 2.1 offline subset rising from 15.7% to 72.4% and GDPval-AA v2.1 at 1620, while Terminal-Bench 4.0 is reported at either 39.2% or 33% across sources, still well below Sonnet 5.5. Haiku 5.5 is the first Haiku model with adjustable effort settings and is positioned as a subagent for Opus 5.5 and Sonnet 5.5; Anthropic also cut Sonnet 5.5 cache-read pricing from $0.20 to $0.10 per million tokens.
IT之家 AI · 🔥 0 - GPT-6 rolls out to all ChatGPT tiers with interactive Intelligent UI
OpenAI began rolling GPT-6 out to ChatGPT on October 7 alongside Intelligent UI, which lets replies include charts, buttons, forms, calculators and other interactive elements, and lets the model start answering while it is still thinking. Plus, Pro, Business and Enterprise users get GPT-6 Sol, while Free and Go users were to receive the less capable GPT-6 Luna starting October 8; Intelligent UI works from Instant through Extra High effort but not at the Pro effort using GPT-6 Astra, and models used by Work and Codex are unchanged. OpenAI reported that GPT-6 Instant's time to begin answering web-search questions averaged 44% shorter than GPT-5.6 Instant, and it also published safety results including a 28% warning-violation rate for GPT-6 Sol and 15.9% for Luna in one test.
ChatGPT Release Notes · 🔥 0 - AI Price Competition Shifts Toward Efficiency and Agent Scale
An analysis argues that the latest AI price competition is being driven not only by lower per-token prices, but also by models completing tasks with fewer tokens and tool calls. It attributes the shift to improvements in inference, caching, reinforcement learning, distillation and engineering, as OpenAI, Anthropic, Xiaomi and SpaceXAI target the growing cost of agent workloads.
创业邦 科技 · 🔥 0
Experience and discussion from the community
Share my GPT-6 Luna experienceAsk about GPT-6 Luna
Nobody has shared their experience with GPT-6 Luna yet.