与「gpt」相关的搜索结果

国内好多人都说Gemini是美国版豆包？我觉得这两者都不是一个量级的吧？Gemini可以混合notebooklm记笔记有多好用，用过都知道，还有深度联网搜索功能用来做深度详细报告，还有canvas模式（这个更是夯爆了好吗？） 而且Gemini的可视化方面，还有审美ui设计方面绝对是世界顶级模型，还可以和Google ai studio混合交叉使用，夯中夯，还有生图功能的Nano Banana模型，在gpt的image2没出来之前一直是最强的图形生成模型吧，不知道为什么会有人拿Gemini跟豆包比较？那些人真的用过Gemini吗？对此表示怀疑🤔🤔🤔

My entire AI stack is now Chinese 🇨🇳 87% cheaper. same revenue swaps by task: 1. reasoning / backend brain Opus 4.8 → Kimi K2.7 benchmark gap: ~8% · price: ~11x cheaper 2. code generation GPT-5.5 → Qwen 3.7 Max benchmark gap: ~18% · price: ~7x cheaper 3. agent loops + tool calling Sonnet 4.7 → GLM 5.2 benchmark gap: ~3% · price: ~5x cheaper on input 4. cheap volume / bulk processing GPT-5.5 mini → MiMo V2.5 benchmark gap: ~6% · price: ~12x cheaper 5. image generation GPT-Image-2 → Wan 2.5 benchmark gap: ~5% · price: ~8x cheaper 6. video generation Sora 2 → Kling 3.0 benchmark gap: roughly equal · price: ~6x cheaper [ result after 30 days: ] operating costs dropped 87%, output quality dropped 4% on average, revenue unchanged the most important that these models will be not banned in a month and i can run them locally nobody will steal my data and i can learn them as i need full article drops tomorrow with: > exact routing logic per task type > the 2 cases where I still pay for American > the migration playbook anyone can copy in a weekend VERY IMPORTANT to get migrated now, while it's not too late

Claude最近实在没法用了，Opus 4.8幻觉问题爆炸，还总是莫名其妙出现一些别人的信息串到我的对话里，整个周末我代码都被幻觉问题改崩了好几次，每次任务做到一半就会被各种带偏，用Codex修到今天才把坑全部修完。 忍无可忍了，重开了ChatGPT的200刀会员，重回Codex怀抱，坐等GPT-5.6。。。 而且有一说一，Codex的远程控制体验真的吊打Claude N个数量级。

K-POP IDOL Day Off Vlog Image: GPT Image 2 Video: Seedance 2.0 on @DomoAI_ ======================= Prompt: 15 seconds, 16:9. Main subject from @ Image1 — lock her completely. Face, skin tone, body, hairstyle, and outfit must remain perfectly identical throughout the entire video. She wears a sleeveless green-and-white striped ribbed knit mock-neck top with a red-and-white striped collar, medium-wash loose straight-leg high-waisted blue jeans, and simple white low-top sneakers. No bag. Long straight dark brown hair with a center part. Korean woman, 20 years old, naturally beautiful with the effortless charm of a modern K-pop idol. The entire video is filmed exclusively as a handheld front-facing smartphone selfie vlog. The phone is always in her own hand. There are absolutely no third-person shots, no external camera, no tripod, no cinematic coverage, and no impossible viewing angles. Every shot is from the perspective of her personally recording herself during a relaxing day off. Modern Seoul residential neighborhood. Quiet apartment complexes, tree-lined sidewalks, small neighborhood park, convenience store, peaceful alleys, warm summer morning light. Everything feels authentic and lived-in. Natural smartphone selfie camera language only. Slight arm movement, subtle handheld shake, natural walking bounce, occasional autofocus adjustment, realistic exposure adaptation, small framing imperfections, slight rolling shutter while walking, natural lens distortion from the front camera, occasional brief glances at the screen to check framing. Crisp modern smartphone image quality with realistic HDR and natural colors. No cinematic grading. No vintage effects. Scene 1: She leaves her apartment building while holding the phone at arm's length. She smiles naturally into the selfie camera and says, "오늘은 쉬는 날이라 그냥 산책 나왔어요." ("It's my day off, so I just came out for a walk.") Scene 2: Still filming herself, she enters a convenience store, opens the refrigerator with one hand while holding the phone with the other, picks up an iced coffee, smiles at the camera and says, "이거 마셔볼게요." ("I'm going to try this.") Scene 3: Holding the phone herself, she walks slowly through a quiet neighborhood park while sipping her coffee. She looks up at the trees, smiles, then back into the camera. "오늘 날씨 너무 좋다." ("The weather is so nice today.") Scene 4: She notices a friendly stray cat beside the sidewalk. Still filming herself, she lowers the phone slightly so both she and the cat remain in frame, crouches down and smiles. "안녕~ 귀엽다." ("Hi~ You're so cute.") Scene 5: Continuing her walk, she talks casually to the selfie camera while brushing her hair behind one ear. "이렇게 평범한 하루가 제일 행복한 것 같아요." ("I think ordinary days like this make me the happiest.") Final scene: While continuing to walk, she smiles directly into the front camera, waves with her free hand and says, "다음 브이로그에서 만나요. 안녕!" ("See you in the next vlog. Bye!") She laughs naturally, lowers the phone toward her side, and the recording ends exactly as if she pressed the stop button herself. Audio is entirely natural. Her live Korean voice, footsteps, soft wind, birds, distant traffic, convenience store ambience, leaves rustling, occasional cat meow. No background music, no narration, no cinematic sound design, no added effects. #DomoAI#

要不怎么还是奥特曼带领的 OpenAI “圆滑”呢，或者说真的在 Open？ 有用户爆料，现在 Codex 会将部分 GPT-5.5 xhigh 会话默默路由到 GPT-5.6-SOL😂 换成 Anthropic ，这事儿应该就不可能发生。

买了日本和A股的机器人ETF。 机器人的爆发已经在路上了 大模型是AI的大脑 机器人是AI的身体 大脑需要建好，需要存储、GPT/CPU 但大脑最终需要身体 身体最终需要去创造价值 未来存储的需求是现在的100倍：机器人、视频、模拟世界/游戏。

Community report: Codex may be quietly routing some gpt-5.5 xhigh sessions to gpt-5.6-sol. Try it in Codex App/CLI: select gpt-5.5, set reasoning to xhigh, then run the Juice test prompt. If it returns 128, you are probably in the gray rollout. Worth testing to see if you got it.

大事！土区ChatGPT Plus 自从涨价后 身边好多朋友土区老账号弹出涨价后的标价，属实猝不及防 一直用土区开Plus就是图划算，这下优势要没？ 重点求证：咱们现有订阅，7月18号前会不会调价？ 专门开打卡追踪，今天第1天 Day01，每天同步真实账号价格， 后续也会持续更新账号扣费、续费规则、避坑办法，持有土区GPT的朋友别错过🤝