Image created with OpenAI gpt-image-1. Image prompt: Single-panel cartoon with loose, hand‑inked lines, bean‑bodied figures, muted flat colors, minimal props, and deadpan humor: Treasure chest labeled “OpenAI” bursts open; curious ideas in lab coats scramble into daylight. Large bold title text centered at top: “OPENAI” Muted colors, flat shading, black ink outlines. 16:9. Caption: “Finally living up to the name.”
“o3’s search abilities are incredible. Can find extremely niche information without a ton of additional context. Just what I would say to a colleague.” / X https://x.com/natolambert/status/1913433658909720880
OpenAI would buy Google’s Chrome, exec testifies at trial | Reuters https://www.reuters.com/sustainability/boards-policy-regulation/google-contemplated-exclusive-gemini-ai-deals-with-android-makers-2025-04-22/
The Washington Post partners with OpenAI on search content – The Washington Post https://www.washingtonpost.com/pr/2025/04/22/washington-post-partners-with-openai-search-content/
“Image gen is now available in the API! We’re launching gpt-image-1, making ChatGPT’s powerful image generation capabilities available to developers worldwide starting today. ✅ More accurate, high fidelity images 🎨 Diverse visual styles ✏️ Precise image editing 🌎 Rich world https://x.com/OpenAIDevs/status/1915097067023900883
“a few things are different about the api version than the chatgpt version: you can control moderation sensitivity with the ‘moderation’ parameter you can also control things like quality vs generation speed, background, output format, etc.” / X https://x.com/sama/status/1915111808983282010
“How does the OpenAI Codex CLI work? Yesterday, OpenAI released a open-sourced Codex a “chat-driven development” CLI. It allows developers to use AI models via API directly in their terminal to perform coding tasks. Unlike a simple chatbot, it can read files, write files (via https://x.com/_philschmid/status/1912870519294091726
“5/ OpenAI just released a 34-page practical guide to building agents, thanks for sharing @Hesamation https://x.com/AtomSilverman/status/1913372919851614533
“OpenAI just released a 34-page practical guide to building agents, here’s 10 things it teaches us: https://x.com/Hesamation/status/1912916069699793006
“Can AI meaningfully help with bioweapons creation? On our new Virology Capabilities Test (VCT), frontier LLMs display the expert-level tacit knowledge needed to troubleshoot wet lab protocols. OpenAI’s o3 now outperforms 94% of expert virologists. https://x.com/DanHendrycks/status/1914696657813561799
“💥 We just launched ChatGPT’s imagegen model in the API! It went viral in ChatGPT, and our API offers even greater parameter control to match diverse developer use cases. Also, we doubled rate limits for o3 and o4-mini for ChatGPT Plus users. Happy Wednesday 🌞” / X https://x.com/kevinweil/status/1915103387592409215
“OpenAI’s o3 and o4-mini scores on the Extended NYT Connections benchmark This benchmark evaluates large language models (LLMs) using 651 NYT Connections puzzles, with additional words included to increase difficulty. The standard NYT Connections benchmark is nearing saturation, https://x.com/rohanpaul_ai/status/1913927366717342166
“We’ve been collaborating closely with developers to understand where image gen can be most useful in the real world. Here are some examples from early adopters across domains like creative tools, consumer apps, enterprise software, and more below. 👇” / X https://x.com/OpenAIDevs/status/1915097072107143322
“Now that AI driven search works pretty well using o3, the implications are pretty big… https://x.com/emollick/status/1914366069160206482
Business guides and resources | OpenAI https://openai.com/business/guides-and-resources/
“i think this is gonna be more like the renaissance than the industrial revolution” / X https://x.com/sama/status/1913320105804730518
“OpenAI’s new o3 model hallucinates significantly more than older releases! The findings come from third-party testing, where the AI was caught lying about actions it never took! OpenAI’s system card also says o3 seems to hallucinate >2x more than o1 https://x.com/rowancheung/status/1914201289988227506
“A potential issue with o3 is that it thinks it is using tools even when it does not, leading to some hallucinations where it assumes work that was implied in the reasoning chain was actually done. You should double check the reasoning trace for complex work to see what it did.” / X https://x.com/emollick/status/1912889596125274283
“”A practical guide to building agents” from @OpenAI It’s a guide for product and engineering teams building their first AI agents, covering: – Use case selection – Agent design patterns – Best practices for safe, reliable deployment – And other foundation you need to start https://x.com/TheTuringPost/status/1913002164475351212
OpenAI Forecasts Revenue Topping $125 Billion in 2029 as Agents, New Products Gain — The Information https://www.theinformation.com/articles/openai-forecasts-revenue-topping-125-billion-2029-agents-new-products-gain
“Turns out, LLMs represent numbers on a helix and use trigonometry to do addition. A new paper reverse engineers addition in models like GPT-J-6B and finds a “Clock” algorithm. Numbers are encoded using sine and cosine terms, then added like angles. https://x.com/LiorOnAI/status/1914334179929530660
“killer story from @srimuppidi: openai has upped its projections even more from last fall. key drivers include agents and “free user monetization.” that second driver could refer to channels including advertising or affiliate fees: https://x.com/steph_palazzolo/status/1915095524677505049
“Deep Research (lightweight version) is now available in the ChatGPT free tier:” / X https://x.com/gdb/status/1915637620731941188
“I’m hearing more and more stories of ChatGPT helping people fix longstanding health issues. We still have a long way to go, but shows how AI is already improving people’s lives in meaningful ways. https://x.com/gdb/status/1914106403574452496
“More evidence that o3 represents a big move forward, this time on ARC-AGI. https://x.com/emollick/status/1914798775840706690
Not For Private Gain https://notforprivategain.org/
“All Plus, Team, and Pro users now have expanded usage of deep research. Enterprise and Edu users will get access the following week, and will have the same usage levels as Team users. The lightweight version of deep research is starting to rollout today to all Free users.” / X https://x.com/OpenAI/status/1915505964142514321
Investigating truthfulness in a pre-release o3 model | Transluce AI https://transluce.org/investigating-o3-truthfulness
“@tomieinlove tens of millions of dollars well spent–you never know” (re saying “please” and “thank you” / X https://x.com/sama/status/1912646035979239430
“deploying fusion to the world is surprisingly gated on high-power semiconductors. the most important place you could design and build those high-power semiconductors (and do other awesome electrical engineering) is helion. please consider: https://x.com/sama/status/1913609769518317929
“🩵 we will work hard to make much more beautiful creations!” / X https://x.com/sama/status/1913076070007677017
“o3 is far more agentic than people realize. Worth playing with a lot more than a typical new model. You can get remarkably complex work out of a single prompt. It just does things. (Of course, that makes checking its work even harder, especially for non-experts.) https://x.com/emollick/status/1913471315807191310
“OpenAI recently released a guide on building agents which contains some misguided takes There’s a lot of FUD, confusion, hype, and noise around agents I wrote a blog on how to think about agent frameworks. Includes: Background Info – What is an agent? – What is hard about https://x.com/hwchase17/status/1914016302261506421
“OpenAI dropped the new o3 and o4-mini reasoner models o3 pushes SOTA performance across coding, math, science, and multimodality, while o4-mini offers fast, cost-efficient performance Both have agentic access to ChatGPT tools and can “think with images” https://x.com/rowancheung/status/1912744116053590087
“The geoguessing power of o3 is a really good sample of its agentic abilities. Between its smart guessing and its ability to zoom into images, to do web searches, and read text, the results can be very freaky. I stripped location info from the photo & prompted “geoguess this” https://x.com/emollick/status/1912726124913623143
“Here is a new open-source IDE to help you build multi-agent systems. It’s like Cursor but specifically for building multi-agent workflows. It’s powered by OpenAI Agents SDK, connects MCP servers, and can integrate into your apps using HTTP or the SDK. https://x.com/omarsar0/status/1915040973601776092
“5/ An MCP server to browse the web like humans! Stagehand by @browserbasehq allows Agents to independently control a browser. Integrating its MCP server with Claude delivers a more reliable alternative to OpenAI Operator. https://x.com/AtomSilverman/status/1912963628313284863
“Breaking: new @OpenAI models shake up the Arena leaderboard🔥 Highlights: – o3 #2 overall, ties Gemini-2.5-Pro at #1 in Style Control, Math, Coding, and Hard Prompts – o4-mini breaks into top 10 and claims #1 in Math, surpassing o1 (!) – GPT-4.1 ranks top-5 in Hard Prompts, https://x.com/lmarena_ai/status/1915078057452573142
“Why OpenAI doesn’t include other models in their own benchmarks: OpenAI-MRCR results with Gemini 2.5 Pro https://x.com/scaling01/status/1913955228442833032
“To start using gpt-image-1, verify your organization: https://x.com/OpenAIDevs/status/1915097084308410816
“Vision models still do much worse than humans in being able to accurately answer questions about images, but progress has been rapid and o3 represents a big jump forward.” / X https://x.com/emollick/status/1914609409172418756
OpenAI has launched the ChatGPT Image Library – ChatGPT – OpenAI Developer Community https://community.openai.com/t/openai-has-launched-the-chatgpt-image-library/1230140
Image generation – OpenAI API https://platform.openai.com/docs/guides/image-generation?image-generation-model=gpt-image-1
“GPT Image Generation Try the latest image generation model from OpenAI that supports contextual image generation and editing. https://x.com/perplexity_ai/status/1915819619333640647
Introducing our latest image generation model in the API | OpenAI https://openai.com/index/image-generation-api/
“Introducing OpenAI o3 and o4-mini—our smartest and most capable models to date. For the first time, our reasoning models can agentically use and combine every tool within ChatGPT, including web search, Python, image analysis, file interpretation, and image generation. https://x.com/OpenAI/status/1912560057100955661
“Announcing the first Codex open source fund grant recipients: ⬩vLLM – inference serving engine @vllm_project ⬩OWASP Nettacker – automated network pentesting @iotscan ⬩Pulumi – infrastructure as code in any language @pulumicorp ⬩Dagster – cloud-native data pipelines @dagster” / X https://x.com/OpenAIDevs/status/1915524612970152376
ChatGPT now lets users create fake images of politicians. We stress-tested it | CBC News https://www.cbc.ca/news/canada/chatgpt-fake-politicians-1.7507039
OpenAI helps spammers plaster 80,000 sites with messages that bypassed filters – Ars Technica https://arstechnica.com/security/2025/04/openais-gpt-helps-spammers-send-blast-of-80000-messages-that-bypassed-filters/
“Elon Musk’s xAI updated Grok with Memory and Studio Memory enables the AI to remember chats for personalized responses, while Studio creates a new window to help users collaborate with it on docs, code, etc. Studio is pretty much like ChatGPT Canvas! https://x.com/adcock_brett/status/1913986949112631554
Flex processing – OpenAI API https://platform.openai.com/docs/guides/flex-processing
Sam Altman steps down as Oklo chair https://www.cnbc.com/2025/04/22/sam-altman-steps-down-as-oklo-chair-freeing-nuclear-company-up-to-work-with-more-ai-companies.html
“imagegen is launched in the openai api! build cool stuff plz” / X https://x.com/sama/status/1915110344894435587
OpenAI seeks to make its upcoming ‘open’ AI model best-in-class | TechCrunch https://techcrunch.com/2025/04/23/openai-seeks-to-make-its-upcoming-open-ai-model-best-in-class/
“We’ve noticed many of you love using deep research, so we’re expanding usage for Plus, Team, and Pro users by introducing a lightweight version of deep research in order to increase current rate limits. We’re also rolling out the lightweight version to Free users.” / X https://x.com/OpenAI/status/1915505959931437178
“chatgpt is great now. i sort of think we should put up the version from december of 2022 so people can see how far we’ve come! would feel so ancient.” / X https://x.com/sama/status/1912919833290916234
“the downside to the memory feature is that there’s no way to “send prompt” – as soon as I realized how powerful it was I put some deliberate effort into building persistent respect & rapport with the models and now my chatgpt experience is different.” / X https://x.com/gallabytes/status/1914910758472978770
“You can now use Codex with any model (not just OpenAI models). Game on if OpenAI buys Windsurf and lets us pick any model! Cursor will 👀.” / X https://x.com/svpino/status/1914439870677967015
“What OpenAI releases used to look like just 1 year ago, with comparisons to other models: https://x.com/scaling01/status/1913957941645889583
OpenAI’s o3: Over-optimization is back and weirder than ever https://www.interconnects.ai/p/openais-o3-over-optimization-is-back
“Although o3 does not have access to a coding tool, it claims it can run code on its own laptop “outside of ChatGPT” and then “copies the numbers into the answer” We found 71 transcripts where o3 made this claim! (3/) https://x.com/TransluceAI/status/1912552050766016838
Analyzing o3 and o4-mini with ARC-AGI https://arcprize.org/blog/analyzing-o3-with-arc-agi
“@nearcyan I feel like the goalpost movement in my tl is in the reverse direction recently, with LLMs solving prompt puzzles and influencers hyperventilating about AGI. The original OpenAI definition is the one I’m sticking with, I’m not sure what people mean by the term anymore.” / X https://x.com/karpathy/status/1913741942221144430
OpenAI o3 and o4-mini System Card https://simonwillison.net/2025/Apr/21/openai-o3-and-o4-mini-system-card/
“We’ve removed flow action credits. In fact, we’ve fully revamped our pricing structure with 1 goal in mind: simplicity. And, we’re extending free GPT 4.1 and o4-mini for another week. Listen to the full announcement below. https://x.com/windsurf_ai/status/1914383253731205148
“gpqa (requires more world knowledge): o3 > o4-mini instruction following/chat/emotional reasoning (stronger base model helps for this): o3 > o4-mini codeforces (it’s simple space but you got to think reallt really hard): o4-mini > o3 aime/math (same as above, more knowledge” / X https://x.com/willdepue/status/1914549086822293916
“we have doubled rate limits for o3 and o4-mini-high for chatgpt plus subscribers. enjoy!” / X https://x.com/sama/status/1915098418504241469
“New Models Grok 3 Beta and o4-mini are now both available in the model selector. https://x.com/perplexity_ai/status/1915819644256129424
“🏆 Detailed breakdowns for how the new OpenAI model family vs other top AI: – o3 #1 in Style Control, Hard Prompts, Coding, Math – o3 & o4-mini both #1 in Math – GPT-4.1 top-5 in Hard Prompts, Math, Longer query https://x.com/lmarena_ai/status/1915078061126725755
“The lightweight version of deep research is powered by a version of OpenAI o4-mini and is nearly as intelligent as the deep research people already know and love, while being significantly cheaper to serve. Responses will typically be shorter while maintaining the depth and https://x.com/OpenAI/status/1915505961500070245
