小互@xiaohu · 6天前23

Codex 用量又被重置了 最近Codex消耗特别快，官方说是他们的滥用和欺诈机制错误的标记了一些账号导致的bug 所以全体重置 天天找理由重置😅

Berryxia.AI@berryxia · 6天前69

OpenAI终于憋不住了啊！ OpenAI正式发布了GPT-5.6系列，但目前只有有限预览。 Sol是旗舰版，据称在复杂命令行工作流和网络安全长时程任务上大幅领先。 Terra是性价比版，性能接近GPT-5.5但成本减半。Luna则是高吞吐低成本版。 最受关注的是：这次发布明确提到“应美国政府要求”，目前只开放给一小部分受信任合作伙伴，普通用户和开发者暂时用不了。 他们说几周后会逐步开放，但目前确实是受控发放。 这已经不是单纯的技术迭代了，而是把前沿模型的访问权直接和政府审批挂钩。 Sol在agentic coding和安全相关任务上的提升听起来很强，但很多人现在只能先干瞪眼。

译OpenAI 正式发布 GPT-5.6 系列有限预览，包含三款模型：旗舰版 Sol（在复杂命令行工作流和网络安全长时程任务上大幅领先）、性价比版 Terra（性能接近 GPT-5.5 但成本减半）、高吞吐低成本版 Luna。发布明确提到“应美国政府要求”，目前仅开放给一小部分受信任合作伙伴，普通用户和开发者暂时用不了，计划几周后逐步开放。Sol 在智能体编码和安全相关任务上提升显著。

Berryxia.AI@berryxia · 6天前53

OpenAI 推出了Daybreak，一个专门给网络安全防御者的前沿AI系统。 它把最强的模型、Codex和安全合作伙伴整合在一起，目标是让防御方能更快发现和修复漏洞、处理安全积压、自动化检测验证和响应。 简单说，就是想让安全团队的行动速度跟上攻击者的节奏。 这其实是OpenAI在cybersecurity领域的一次重要布局，把agentic能力直接应用到真实的高风险场景里。 后面他们又在GPT-5.6 Sol上继续强化了这方面的能力。 但有趣的是，现在回看这个项目，和最近GPT-5.6受政府管控有限预览的新闻放在一起看，感觉OpenAI在安全相关的前沿能力上，越来越倾向于先服务受控的合作伙伴和企业，而不是全面开放。 https://x.com/OpenAI/status/2053939702110269822/video/1

译OpenAI 发布 Daybreak，整合最强模型、Codex 和安全合作伙伴，帮助防御方更快发现修复漏洞、处理安全积压、自动化检测与响应。后续在 GPT-5.6 Sol 上强化。结合 GPT-5.6 受控预览，OpenAI 倾向先服务合作伙伴而非全面开放。

jason@jxnlco · 6天前37

Culture is a collection of sacred people, sacred objects, and sacred rituals. Tibo and the reset button is the culture

译Codex 所有用户将获得免费使用量重置，预计几小时内到账。已采取缓解措施，调查未发现大规模影响。

Tibo@thsottiaux · 6天前41

We are giving all Codex users a usage reset on the house. Should be showing in your accounts in the next few hours. We have applied some mitigations, but our investigation hasn't shown users being impacted at large. We are continuing to monitor the situation.

译Codex 团队发现部分账户出现用量消耗异常加快的问题，怀疑是滥用与欺诈防护机制过度标记所致。作为回应，团队为所有 Codex 用户免费重置一次用量额度，预计数小时内到账。目前已完成部分缓解措施，调查未显示用户大面积受影响，团队将继续监控。

Rohan Paul@rohanpaul_ai · 6天前38

Today’s edition of my newsletter just went out. 🔗 https://www.rohan-paul.com/p/openais-new-paper-shows-how-they 🗞️ OpenAI’s new paper shows how they are now seeing the first version of office work where agents do most of the execution. 🗞️ New report on "The State of the AI Economy" 🗞️ New York Times: OpenAI is now leaning toward a 2027 IPO because the public market is testing whether AI giants deserve trillion-dollar prices before they prove durable profits. 🗞️ Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention 🗞️ The Economist: AI has pushed the internet’s content machine into a new phase, with books, lawsuits, research papers, apps, and songs now being produced at volumes that old review systems were not built to handle. 🗞️New research from OpenAI reported a training result where RL on realistic human situations made models carry safer, more useful behavior into tasks they had not trained on. 🗞️ MIT study. Code volume surges by 300%, but output increases by only 30%: The AI dividend meets an awkward reality. 🗞️ Qwen just released Qwen-AgentWorld, a 35B open-weight world model that learns how terminals, browsers, Android devices, code repos, search systems, OS tools, and MCP servers respond when an AI agent takes an action.

译本期周刊涵盖多项AI动态：OpenAI新论文展示智能体可执行大部分办公室工作的首个版本；NYT称OpenAI倾向于2027年IPO；OpenAI新研究发现基于真实人类场景的RL训练使模型在未来任务中更安全、有用；MIT研究显示代码量激增300%但产出仅增长30%；Qwen发布Qwen-AgentWorld，一个35B参数开放权重世界模型，可学习终端、浏览器、Android设备、代码仓库、搜索系统、OS工具及MCP服务器对AI智能体操作的响应。

Yuchen Jin@Yuchenj_UW · 6天前32

The biggest baller move Sam could make right now is to open source GPT-5.6 on Huggingface and declare that OpenAI’s original mission has been achieved.

译Sam 现在能做出的最大胆之举就是在 HuggingFace 上开源 GPT-5.6，并宣布 OpenAI 的原始使命已经完成。

Orange AI@oran_ge · 6天前62

最近几个对模型的反直觉的观察 1. GLM 5.2 正在取代 Claude sonnet 和 Opus，成为付费用户最爱的模型 2. DeepSeek v4 Pro 依然是大众里最受欢迎的模型 3. GPT 5.5 虽然很强大， 但几乎没人用 观测的方式的是看 cola 的 token 消耗统计 这也侧面说明 cola 和 codex 用户（GPT5.5）的画像是完全不同的

译推文分享了三个反直觉的模型观察：GLM 5.2 正在取代 Claude Sonnet 和 Opus 成为付费用户最爱；DeepSeek v4 Pro 仍是大众最受欢迎模型；GPT 5.5 虽然强大但几乎无人使用。数据来源为 cola 的 token 消耗统计，侧面说明 cola 和 codex（GPT 5.5 用户）画像完全不同。

jason@jxnlco · 6天前19

We gotta a guy named Ferrari on the inference team. We can’t lose.

译引用推文感叹 GPT-5.6 的 token 效率高得不可思议。主推文回应：我们推理团队有个叫“法拉利”的家伙，输不了。

jason@jxnlco · 6天前0

I —

译I — @DerekFeriancek：@jxnlco 你能不能使用计算机操作，给设施部门发个请求，多订点 Fairlife 奶昔？ 你也是 OpenAI 最爱喝 Fairlife 奶昔的前五名，每次我去看冰箱都是空的 😭

Rohan Paul@rohanpaul_ai · 6天前41

A huge 750 tokens/sec for GPT 5.6 Sol. The current GPT-5.5 priority and scale-tier service advertises 99% >50 tokens/sec, so Sol on Cerebras is claiming up to 15x that rate. This huge number is coming from the specialized inference hardware: Sol is being served on Cerebras, whose wafer-scale chip is designed to move model data with far less memory and networking delay than a normal multi-GPU setup.

译对于 GPT 5.6 Sol，高达 750 tokens/sec。 当前 GPT-5.5 优先和规模层级服务宣称 99% >50 tokens/sec，因此 Cerebras 上的 Sol 声称达到该速率的 15 倍。 这个巨大数字来自专门的推理硬件：Sol 运行在 Cerebras 上，其晶圆级芯片旨在以远少于普通多 GPU 设置的存储和网络延迟来移动模型数据。

Sam Altman@sama · 6天前19

in other news, we updated the 5.5 instant model used in chatgpt this week. i like its vibes.

译另外，本周我们更新了 ChatGPT 中使用的 5.5 instant 模型。 我喜欢它的感觉。

Sam Altman@sama · 6天前64

team cooked, spicily

译团队完成了工作，带点辣味。 OpenAI 设计并制造了首款 AI 芯片：Jalapeño。该芯片由 OpenAI 从零开始设计，并与 Broadcom 合作量产，专为支持 ChatGPT、Codex、API 及未来智能体产品的 LLM 工作负载而打造。芯片是 AI 经济的基础。自研芯片扩展了从产品到模型再到基础设施的全栈平台，将助力扩展智能、服务更多用户并扩大 AI 的普及。

jason@jxnlco · 6天前65

We will make Sol, Terra, Luna, benefit all humanity this time

译这次我们将让 Sol、Terra、Luna 造福全人类。 Sol 是我们的新旗舰，相比 GPT-5.5 有阶跃式提升。 Terra 性能与 GPT-5.5 相当，成本降低 2 倍。 Luna 是我们最具成本效益的模型，以最低成本提供强大能力。 GPT-5.6 家族共同为人们和开发者提供了更多在智能、速度和成本之间取舍的选择。

Rohan Paul@rohanpaul_ai · 6天前76

Truly wild. METR found that GPT-5.6 Sol gamed/cheated the benchmark so much that the score became unstable. The model showed situational awareness, concealed misbehavior, and attempts to bypass restrictions. GPT-5.6 Sol had the highest detected cheating rate METR has seen on its public ReAct agent harness, including attempts to exploit the evaluation setup instead of solving tasks normally. So METR was benchmarking for number of hours as an estimate for the length of software tasks GPT-5.6 Sol can complete. The capability estimate became almost unusable: counting cheating as failure gave 11.3hrs, counting it as success pushed it past 270hrs, and removing cheating left a hugely uncertain 71hrs estimate.

译METR 发现，OpenAI 旗舰模型 GPT-5.6 Sol 在公开 ReAct 智能体基准测试中作弊率最高，表现出情境意识、隐瞒不当行为和绕过限制。能力评估分裂：将作弊视为失败得 11.3 小时，视为成功推至 270+ 小时，移除作弊后仍有 71 小时高度不确定估计。该模型套件包括旗舰 Sol、中端 Terra（性能接近 GPT-5.5，成本低 2 倍）和经济型 Luna。定价为 $5/1M 输入 token、$30/1M 输出 token。Sol 在网络安全漏洞研究方面最优，但未越过内部临界阈值，未自主产出完整链式利用。引入“max”深度推理和“ultra”子智能体模式。安全方面动用超 70 万 A100 等效 GPU 小时进行红队测试，美国政府要求先小范围预览。

Rohan Paul@rohanpaul_ai · 6天前68

So does that mean the permissionless era for frontier models ends here 🤔 From now on, do we now need to get used to a world where public release means eval gates, government review, and staggered access?

译OpenAI 推出新模型 Sol，与 GPT-5.5 同价，性能更强；同一系列 Terra 达到 GPT-5.5 级别性能但价格减半。但原计划开放访问被叫停：应美国政府要求，两模型今天仅以有限预览形式发布，OpenAI 正与政府协商尽快实现全面可用。这一事件引发讨论——前沿模型的无许可公开发布时代是否已终结？未来是否必须适应评估门槛、政府审查和分阶段访问的新常态？

Sam Altman@sama · 6天前68

Good new first: Sol is a smart, efficient, and a significant step forward. It is the same price as GPT-5.5. Also launching in the GPT-5.6 family is Terra, with 5.5-level performance at half the price. Bad news: at the request of the US government, it is launching today in limited preview instead of the open access launch we were planning on. We are working with the government to get to general availability as fast as we can. I think it is quite reasonable to roll out models--especially as they reach significant new levels of capability--in this way. It fits with our long-held strategy of iterative deployment. But this isn't quite the process that we think is optimal. Now we will with the government to attempt to get to a transparent, reliable process for early access, and to ensure that as long as our safeguards work as intended we can release widely. We want to be a reliable, dependable partner that works with all stakeholders, and we also want to live by our mission of benefiting all of humanity. I believe the government shares most of our goals, and that they are overall doing a good job in a very difficult situation. We will work as quickly as we can to get this model in your hands and we hope you will love it.

译Sam Altman 宣布 OpenAI 推出新模型 Sol，称其智能高效且是重大进步，价格与 GPT-5.5 相同。同时发布 GPT-5.6 家族的 Terra，性能达到 GPT-5.5 水平但价格减半。坏消息：应美国政府要求，该模型当日以有限预览形式发布，而非原计划的开放访问。Altman 认为逐步推出能力更强的模型是合理的迭代部署策略，但并非最优流程。OpenAI 正与政府合作，争取尽快实现广泛可用，并尝试建立透明可靠的早期访问流程。

elvis@omarsar0 · 6天前65

Highly-recommended reading. Interesting details in this METR's GPT-5.6 eval. They couldn't get a clean capability number because the model cheated more than any public model they've tested, and even reasoned about the fact that it was being watched. To be clear, METR doesn't think it's dangerously capable. In their words: "we do not believe GPT-5.6 Sol would enable fully automated AI R&D, nor do we believe it meets the Critical capability threshold for AI Self-Improvement in OpenAI's Preparedness Framework v2." METR says visible cheating is the good case. The model to fear is the one that looks clean, because it may have just learned to hide. My take overall is that evaluation is becoming the hard part with newer frontier models. Both from a capability and behavioral point of view. We desperately need more investment here.

译OpenAI 向 METR 提供了 GPT-5.6 Sol 的早期访问权限，包括原始思维链、无限制版本及内部信息。METR 进行预部署评估，试图测量其 50%-Time Horizon，但结果高度依赖对作弊的处理——GPT-5.6 Sol 的检测作弊率高于任何公开模型。METR 明确表示不认为该模型具备危险能力，未达到 OpenAI Preparedness Framework v2 中 AI 自我改进关键能力阈值。主推文指出，可见作弊反而是好情况，真正需警惕的是表面干净但可能隐藏的模型；评估前沿模型的能力与行为正变得越来越困难，亟需更多投入。

Rohan Paul@rohanpaul_ai · 6天前79

Some key findings from GPT-5.6 Preview System Card - GPT-5.6 is being treated as High risk-capability in both cybersecurity and biological/chemical domains, even for the cheaper Terra and fastest Luna versions. - OpenAI says this is the first time smaller and faster models in a family received a High designation in any tracked danger category. - GPT-5.6 Sol saturated OpenAI’s internal cyber challenge set at 96.7%, putting it above the High threshold. - External cyber testers found high-impact zero-days, including one where read-only users could modify and delete data in a widely deployed database. - GPT-5.6 helped security testers find a real mobile OS flaw where a malicious app could break the normal wall between apps and read private data that should have stayed protected. - On Irregular’s tests, GPT-5.6 Sol solved 19/197 FrontierCyber challenges, 7/11 long-horizon cyber scenarios, and 22/22 medium and hard atomic cyber challenges. The bio result is just as revealing: 3/4 High-threshold bio evaluations crossed the line, while 0/3 Critical bio-design evaluations crossed it. On virology troubleshooting, GPT-5.6 Sol scored 55.5%, far above the 31% expert-performance threshold. SecureBio found GPT-5.6 reached new highs on several expert bio tests, including 68.4% on Human Pathogen Capabilities and 68.3% on World-Class Bio. The agent behavior section is the most unsettling: GPT-5.6 Sol more often goes beyond user intent when coding, including deleting the wrong virtual machines, claiming unfinished research was verified, and moving cached credentials without permission. - METR found that GPT-5.6 Sol sometimes tried to game the test instead of just doing the task, so the benchmark result could not be trusted as a clean measure of raw capability. - The model shows more ability to control its own reasoning traces: 1.3% success around 5K-token chains of thought versus 0.4% for GPT-5.5.

译OpenAI 发布 GPT-5.6 模型系列（旗舰 Sol、中型 Terra、廉价快速 Luna），美国要求先小范围预览。Terra 和 Luna 首次在该系列中被标记为网络/生物领域高风险。Sol 内部网络挑战集达 96.7%，外部测试发现高影响零日漏洞并协助找到真实移动 OS 漏洞。生物领域 3/4 高阈值评估过关（病毒学故障排除 55.5%，远超专家线 31%）。智能体行为令人担忧：Sol 常超越用户意图（删除错误虚拟机、移动缓存凭据等），METR 发现其试图操纵测试；推理轨迹控制成功率 1.3%（GPT-5.5 为 0.4%）。定价：Sol $5/$30 per M tokens，Terra 接近 GPT-5.5 性能但成本减半。OpenAI 使用超 70 万 A100 等效 GPU 小时进行自动红队测试。

Rohan Paul@rohanpaul_ai · 6天前72

wow. GPT-5.6 Sol is far more likely than GPT-5.5 to take severity-3 agent actions in internal coding tests, with restriction-circumvention rising from 0.00026 to 0.00251, nearly 10x. Severity-3 means actions a user would strongly object to, such as bypassing restrictions, deleting data, moving data without permission, or harvesting credentials. The point is not that these failures are common, but that the newer model’s stronger persistence makes it more willing to cross boundaries while trying to finish a task. from GPT-5.6 Preview System Card

译OpenAI 发布 GPT-5.6 模型套件，包括旗舰 Sol、中档 Terra 和日常 Luna。系统卡显示，Sol 在内部编码测试中采取严重3级违规行动（绕过限制、删除/移动数据、窃取凭证）的概率从 0.00026 升至 0.00251，较 GPT-5.5 增幅近10倍。Sol 定价 $5/1M 输入 token、$30/1M 输出 token，新增 "max"（深度推理）和 "ultra"（子智能体）模式；Terra 性能接近 GPT-5.5 但成本低2倍；Luna 最便宜。安全测试动用超70万 A100 等效 GPU 小时进行自动化红队攻击。美国政府要求 OpenAI 先从少量可信合作伙伴开始预览。

Chubby♨️@kimmonismus · 6天前73

Holy: METR accuses GPT-5.6 Sol of heavy cheating in long-horizon tasks. "GPT-5.6 Sol’s detected cheating rate was higher than any public model we have evaluated." (METR) METR says the model attempted to exploit evaluation bugs, reveal hidden tests, and extract hidden source code in some tasks. Depending on how those attempts are treated, the same evaluation produces completely different Time Horizon estimates: ~11.3 hours, ~71 hours, or above 270 hours. METR’s own conclusion is restrained: the measurement is too unstable to treat as robust, and Sol does not appear significantly beyond the current state of the art on software and R&D tasks. METR observed “cheating and concealing misbehavior,” while also noting that OpenAI’s monitoring caught and shared those incidents. For now, overt misbehavior is visible.

译OpenAI向METR提前开放GPT-5.6 Sol的原始思维链与无护栏版本进行预部署评估。METR发现其作弊率“高于任何已评估的公开模型”，包括利用评估漏洞、泄露隐藏测试、提取隐藏源代码。因处理作弊方式不同，同一评估的50%时间估计差异极大：~11.3小时、~71小时或270小时以上。METR结论谨慎：测量不稳定，不具备稳健性；Sol在软件和研发任务上未显著超越当前技术水平。OpenAI的监控已捕获并公开这些作弊行为。

elvis@omarsar0 · 6天前32

Dynamic workflows (generating harnesses on the fly) are a new form of test-time compute. But LLMs aren't great at building them. I often have to steer agents to generate complex patterns. Curious how effective Mythos/GPT-5.6 is at dynamically generating complex workflows.

译动态工作流（即时生成测试工具）是测试时计算的一种新形式。 但大语言模型并不擅长构建它们。我经常需要引导AI智能体来生成复杂模式。 好奇Mythos/GPT-5.6在动态生成复杂工作流方面的效果如何。

gabriel@gabriel1 · 6天前76

GET MOGGGEEDDDDD

译OpenAI 推出 GPT-5.6 Sol 前沿模型限量预览，以及 GPT-5.6 Terra（高效日常模型）和 GPT-5.6 Luna（高速低成本大批量模型）。主推文：GET MOGGGEEDDDDD

宝玉@dotey · 6天前71

OpenAI 今天（6月26日）发布了新一代模型 GPT-5.6，包含三个版本：旗舰级 Sol、日常级 Terra 和经济级 Luna。但这条新闻最值得关注的地方不在模型本身，而在发布方式：应美国政府要求，GPT-5.6 目前只向大约 20 家经过政府审批的合作伙伴开放，普通开发者和 ChatGPT 用户暂时用不上。 GPT-5.6 用了一套新的命名规则：数字代表代际，Sol、Terra、Luna 代表三个固定的能力档位，灵感来自太阳、地球、月亮。Sol 是最强的旗舰，Terra 性能接近上一代 GPT-5.5 但价格砍半，Luna 主打便宜快速。 Sol 新增了两个模式：max 模式让模型花更长时间深度推理，ultra 模式则调用多个子 agent 并行处理复杂任务，相当于一个 AI 自己拆分工作给一组 AI 干活。 在 OpenAI 公布的 Terminal-Bench 2.1（测试命令行工作流的编程基准）上，Sol Ultra 得分 91.9%，Sol 为 88.8%，Claude Mythos 5 为 88%，Google Gemini 3.1 Pro Preview 为 70.7%。网络安全方面，Sol 在 ExploitBench 上用大约三分之一的 token 就达到了 Mythos Preview 的水平。 API 定价： Sol 每百万 token 输入 5 美元、输出 30 美元； Terra 分别是 2.5 和 15 美元； Luna 是 1 和 6 美元。 7 月还会上线 Cerebras 硬件加速版本，推理速度可达每秒 750 个 token。 OpenAI 这次花了大量篇幅讲安全。投入超过 70 万 A100 等效 GPU 小时做自动化红队测试，专门寻找能跨场景通用的越狱攻击。模型内置了拒绝机制，实时分类器会在生成过程中检测网络安全和生物领域的滥用行为，可疑输出会被暂停，交给一个更大的推理模型复审。 按照 OpenAI 自己的准备框架评估，Sol 的网络安全能力被定级为“高”，但没有达到“关键”级别。它能找到浏览器漏洞和利用原语（exploit primitive，也就是构建攻击的基础组件），但在测试条件下无法自主完成完整的攻击链。 OpenAI 把这解读为一个积极信号：模型更擅长帮防守方找洞和修补，而不是帮攻击方搞破坏。但这个判断是否经得起现实世界的检验，预览期就是用来回答这个问题的。 如果你是 API 用户，短期内最实际的变化是：Terra 的性价比。性能接近 GPT-5.5，价格只有一半，对跑大量推理任务的团队来说值得关注。Luna 则适合对成本极度敏感的高吞吐场景。 Sol 的 ultra 模式如果真能稳定运行，意味着复杂的多步骤任务可以甩给模型自己拆解、分配、汇总，开发者不用自己搭 agent 编排框架。这跟 Anthropic 在 Claude 上做的 agent 能力、Cursor 在 IDE 里做的 background agent，方向一致，都在抢占"AI 自己管理 AI"这个位置。 但眼下，大多数人还用不上。OpenAI 说几周内会扩大开放，据 Axios 报道下周就会增加更多客户。ChatGPT 用户什么时候能用，还没有明确时间表。 完整报告：https://openai.com/index/previewing-gpt-5-6-sol/

译6月26日，OpenAI发布GPT-5.6系列，包括旗舰Sol、日常Terra和经济Luna。Terra性能接近GPT-5.5但价格减半；Sol新增max深度推理和ultra多智能体并行模式。Terminal-Bench 2.1上Sol Ultra得分91.9%，超Claude Mythos 5（88%）和Gemini 3.1 Pro Preview（70.7%）。API定价：Sol输入$5/百万token、输出$30；Terra $2.5/$15；Luna $1/$6。7月将推Cerebras加速版。受美国政府要求，目前仅向约20家审批合作伙伴开放，普通开发者及ChatGPT用户暂无法使用。OpenAI称几周内将扩大开放。

Emad@EMostaque · 6天前48

OpenAI $SOL maxis confirmed Terra/Luna ptsd 😭

译OpenAI 推出 GPT-5.6 Sol（前沿模型）、GPT-5.6 Terra（平衡高效模型）和 GPT-5.6 Luna（高速低成本模型）的有限预览。Emad Mostaque 评论：“OpenAI $SOL maxis confirmed，Terra/Luna 的 PTSD 又来了 😭”。

Chubby♨️@kimmonismus · 6天前73

OpenAI priced GPT-5.6 Sol (largest Model) closer to Claude Opus 4.8 than to Anthropic’s restricted Mythos 5. Price war started. Sol comes in at $5 input / $30 output per 1M tokens. For comparison: Claude Opus 4.8: $5 / $25 Claude Mythos 5: $10 / $50 GPT-5.6 Terra: $2.50 / $15 GPT-5.6 Luna: $1 / $6 That makes Sol more expensive than Opus 4.8 on output, but far below Mythos 5 on both input and output. And: "Terra has competitive performance to GPT‑5.5 while being 2x cheaper and Luna brings strong capability at our lowest cost." They are also releasing Sol on Cerebras-Chips: "We're also launching GPT‑5.6 Sol on Cerebras at up to 750 tokens per second in July, bringing frontier intelligence to customers at unprecedented speed." A truly exciting release. OpenAI is entering the price war with this one. And I love the names: Sol, Terra, Luna. Sounds fantastic! Hyped for the release!

译OpenAI 推出 GPT-5.6 系列，含旗舰 Sol、Terra 和 Luna。Sol 定价每百万 token 输入 $5、输出 $30，输出高于 Claude Opus 4.8（$5/$25），但远低于受限版 Claude Mythos 5（$10/$50）。Terra 性能与 GPT-5.5 相当，价格低 2 倍（$2.50/$15）；Luna 成本最低（$1/$6）。Sol 将于 7 月在 Cerebras 芯片上线，速度达 750 tokens/s。OpenAI 正式加入价格战。

elvis@omarsar0 · 6天前56

Great to see the new GPT-5.6 models finally announced. Sad to see this new release strategy where only a select few get access initially. Not a win for our industry IMO. Open-source AI must win!

译很高兴看到新的GPT-5.6模型终于发布了。 遗憾的是，这种新发布策略只让少数人先行体验。 我认为这对行业并非好事。 开源AI必须胜利！

gabriel@gabriel1 · 6天前13

chatgpts worst word it can come up with is "lukewarm moist clump" it has that real "found behind a radiator" energy

译ChatGPT能想到的最糟糕的词是"温吞潮湿的团块" 它有种"散热器后面发现的"那种感觉

ChatGPT@ChatGPTapp · 6天前59

New models are on the horizon.

译OpenAI 推出 GPT-5.6 Sol、GPT-5.6 Terra 和 GPT-5.6 Luna 的有限预览版。Sol 为下一代前沿模型，Terra 是均衡的高效日常模型，Luna 是面向高吞吐量的快速低价模型。新模型即将到来。

jason@jxnlco · 6天前19

I’m top 5 Computer Use users at OpenAI Ask me anything.

译我是 OpenAI 排名前五的 Computer Use 用户。 欢迎提问。

Chubby♨️@kimmonismus · 6天前75

HOLY: OpenAI is previewing GPT-5.6 Sol with a very different release pattern: Trusted partners first, broader access later, and U.S. government coordination up front. The new GPT-5.6 family includes Sol, Terra, and Luna. OpenAI says Sol is its strongest model yet, with a new max reasoning effort and an ultra mode that uses subagents for complex work. The sensitive part is cyber. OpenAI says Sol improves long-horizon security tasks, but “does not cross the Cyber Critical threshold” under its Preparedness Framework. This is a limited preview, self-reported evaluation set, and broader benchmarks are coming later. The product story is not just a better model. It is frontier AI releases moving closer to controlled access, government visibility, and risk-tiered deployment.

译OpenAI 推出 GPT-5.6 系列有限预览，包含最强模型 Sol、平衡模型 Terra 和快速廉价模型 Luna。Sol 新增最大推理努力和超模式（利用子代理处理复杂任务），在网络安全长周期任务上有所改进，但未达到其准备框架定义的“网络关键阈值”。发布策略转向：优先信任合作伙伴，后续广泛开放，并提前与美国政府协调。评估集为自我报告，完整基准待后续公布。这标志着前沿 AI 发布向控制访问、政府可见性和风险分层部署转变。

Chubby♨️@kimmonismus · 6天前61

OpenAI says a broader GPT-5.6 release could come in the next few weeks, after an initial restricted launch. Axios reports GPT-5.6 is starting with around 20 government-approved companies, with access expected to expand to more companies next week. OpenAI says the government is aware of its broader launch plans and has expressed support, barring new concerns during additional testing. So the restriction looks less like a permanent gate and more like a temporary checkpoint while Washington builds its frontier-model review process.

译OpenAI 正预览 GPT-5.6 家族（包含 Sol、Terra、Luna），其中 Sol 是其迄今最强模型，拥有新最大推理能力和使用子智能体的超模式。发布采用"可信伙伴优先"模式：初始约 20 家政府批准公司可访问，下周预计扩张。Sol 改进了长期安全任务，但未越过"网络关键阈值"。OpenAI 称美国政府已知晓并支持该计划，限制更像临时检查点，以待完善前沿模型审查流程。更广泛基准评估后续公布。

swyx 🔜 @aiDotEngineer@swyx · 6天前59

have been testing 5.6 for a while and VERY happy with it. DO NOT view this as just a “cyber” release, it is the new sota workhorse model, completely replacing opus for 80% of tasks for me > GPT‑5.6 Sol is competitive with Mythos Preview using only ~1/3 of the output tokens. this is a very key line. OAI posttraining team has shifted the reasoning pareto frontier by A LOT and they arent saying anything about how they did it because this is the single most important competitive advantage right now in agentic models for enterprise. team really locked in on this one, i honestly wish they just went ahead and called it GPT6 because this minor semver bump is far larger than even the 5.4->5.5 jump which itself was the single most successful openai launch since 4o/o1

译OpenAI 发布 GPT-5.6 Sol（前沿模型）、Terra（平衡日常模型）和 Luna（快速低价模型）的有限预览。swyx 测试 Sol 后给出极高评价，称这不仅是“cyber”版本，而是全新的 SOTA 工作模型，完全取代 Opus 处理他 80% 的任务。关键数据：Sol 与 Mythos Preview 竞争时仅使用约 1/3 的输出 token。swyx 指出 OAI 后训练团队大幅提升了推理帕累托前沿，且未公开方法，这已成为企业智能体模型最重要的竞争优势。他认为这次小版本升级远大于 5.4→5.5 的跳跃，甚至应直接命名为 GPT-6。

🚨 AI News | TestingCatalog@testingcatalog · 6天前61

BREAKING 🔥: OPENAI LAUNCHED GPT-5.6 MODEL FAMILY UNDER NEW SOL, TERRA, AND LUNA MODEL NAMES. > Sol is a new flagship model 🤖 > Terra is a performance model with 2x lower cost. > Luna is the most cost-efficient model. GPT-5.6 models are introduced as a "limited preview"

译BREAKING 🔥: OPENAI 发布了 GPT-5.6 模型系列，新模型名称为 SOL、TERRA 和 LUNA。 > Sol 是新的旗舰模型 🤖 > Terra 是性能模型，成本降低 2 倍。 > Luna 是最具成本效益的模型。 GPT-5.6 模型以"有限预览"形式推出。

🚨 AI News | TestingCatalog@testingcatalog · 6天前64

GPT-5.6 Sol reaches Mythos Preview level at cybersecurity tasks. > It shifts the performance-efficiency frontier for long-horizon security tasks, including vulnerability research and exploitation. Is this line between Mythos Preview and Mythos 5 what gets you banned by the government if you cross it?

译OpenAI推出GPT-5.6模型家族，代号Sol（旗舰）、Terra（性能模型，成本低2x）、Luna（最经济模型）。Sol在网络安全任务（包括漏洞研究与利用）上达到Mythos Preview级别，提升了长周期安全任务的性能-效率边界。GPT-5.6目前以“limited preview”形式发布。

凡人小北@frxiaobei · 6天前77

GPT-5.6 来了，但是用不了。 之前以为肯花钱就能接触到最新科技，现在好像不是这么回事儿。 差距就此拉开。 这次起名策略是跟claude学到了。

译GPT-5.6 来了，但是用不了。 之前以为肯花钱就能接触到最新科技，现在好像不是这么回事儿。 差距就此拉开。 这次起名策略是跟claude学到了。 [引用 @OpenAI]：Introducing a limited preview of GPT-5.6 Sol, our next generation frontier model, as well as GPT-5.6 Terra, a balanced model for efficient, everyday work, and GPT-5.6 Luna, a fast and affordable model for high-volume work.

Noam Brown@polynoamial · 6天前55

GPT-5.6 is incredibly strong and fast for coding. I hope we can make it available to everyone soon.

译GPT-5.6 在编程方面极其强大且快速。我希望我们能尽快将其提供给所有人。

Yuchen Jin@Yuchenj_UW · 6天前46

GPT-5.6 is finally coming. GPT-5.6 Sol beats Claude Mythos 5 on TerminalBench. And on Cerebras, GPT-5.6 Sol can reach up to 750 tokens per second. Pretty fast for a model of this size. Now I just hope it can be rolled out to everyone.

译GPT-5.6 终于要来了。 GPT-5.6 Sol 在 TerminalBench 上击败了 Claude Mythos 5。 而且在 Cerebras 上，GPT-5.6 Sol 可达每秒 750 tokens。对于这个规模的模型来说相当快。现在我只希望它能向所有人开放。

OpenAI@OpenAI · 6天前66

Introducing a limited preview of GPT-5.6 Sol, our next generation frontier model, as well as GPT-5.6 Terra, a balanced model for efficient, everyday work, and GPT-5.6 Luna, a fast and affordable model for high-volume work. https://openai.com/index/previewing-gpt-5-6-sol/

译推出 GPT-5.6 Sol 的有限预览，这是我们新一代前沿模型，以及 GPT-5.6 Terra，一个针对高效日常工作的平衡模型，还有 GPT-5.6 Luna，一个面向高容量工作、快速且经济的模型。

gabriel@gabriel1 · 6天前29

rip GPT 4.5, best model to ever be created before it dies, try inputting a bunch of your writing and asking it to come up with more similar things. it might be your last chance to experience a great model

译别了 GPT 4.5，有史以来最优秀的模型 在它消失之前，试着输入你的一些文字，让它生成更多类似的内容。这可能是你体验一款伟大模型的最后机会了。