Gabby Ong (@hellogabbo) profile picture

human (views my own) google deepmind ex startup, figma, AWS

Singapore
Joined April 2016
Photo by Gabby Ong (@hellogabbo)Photo by Gabby Ong (@hellogabbo)Photo by Gabby Ong (@hellogabbo)Photo by Gabby Ong (@hellogabbo)Photo by Gabby Ong (@hellogabbo)Photo by Gabby Ong (@hellogabbo)
New day, new ship! Create worlds in realtime with Genie 4. Start with a single text description or image to set the scene. Then give simple live prompts to shape events, camera angles, and sound as you explore. deepmind.google/models/genie…
Shipping Genie 4, now available to developers! You can try it in the Gemini API and AI Studio. Generate worlds at 24fps with native audio, based on text or image, and then steer it live with WASD
5
2
24
1,082
Below Neon 3D modeled a photorealistic lab simulation in Antigravity via Blender MCP 🔥
352
640
140
14,081
700,466
Gabby Ong retweeted
Gemini 4 Neon from @GoogleDeepMind is out. Google says Gemini 4 Carbon quality but way faster, so we timed it in a voice agent. Neon started talking about 5x sooner than Carbon. Yeah, that’s faster. Give it try on LiveKit Inference today. docs.livekit.io/agents/model…
5
3
25
2,084
"launching periodically" means dropping a new element every 2 days. Neon hits the sweet spot - fast, smart, literally a 10 - excited for ya'll to try it!
Introducing Gemini 4 Neon! Pushing the frontier across domains and punching above its model class in a meaningful way. To me, Neon feels like the right balance of intelligence, speed, and usability. Available now across @GoogleAIStudio, the API, and more!
11
1
7
210
19,876
Gabby Ong retweeted
Google DeepMind's Gemini 4 Neon debuts at #2 on the Vals Index (70.9%), 0.4 points behind Claude Fable 5.2 at less than half the cost per task. It also takes #1 on our Finance and Legal Indexes, Tax Agent and Legal Research.
19
58
16
589
47,905
Gabby Ong retweeted
What a week at @aiDotEngineer! S/o to @swyx for making it happen, and thanks to everyone who came by to say hi and celebrate the week of Gemini. 🚢
8
4
1
54
6,835
Gabby Ong retweeted
it's quite a good model, give it a try!
Gemini 4 Carbon is now available in Cursor! At high effort, it's the top model on CursorBench at 60.4%
131
82
13
3,392
237,895
Give Gemini 4 Carbon a spin on OpenRouter!
Gemini 4 Carbon from @GoogleDeepMind is now live on OpenRouter, via @googlecloud. Built for long-horizon agentic coding and cybersecurity. Google reports big gains over Gemini 4 Argon on exploiting vulnerabilities, agentic terminal tasks, and long-context reasoning. Introductory pricing is 50% off through Nov 30: $2/M input and $10/M output.
5
1
14
973
Gemini 4 Carbon from @GoogleDeepMind is now live on OpenRouter, via @googlecloud. Built for long-horizon agentic coding and cybersecurity. Google reports big gains over Gemini 4 Argon on exploiting vulnerabilities, agentic terminal tasks, and long-context reasoning. Introductory pricing is 50% off through Nov 30: $2/M input and $10/M output.
Gemini 4 Carbon is rolling out today. It’s built for sustained, long-horizon tasks, reasoning deeply through complex software engineering, cybersecurity, and scientific discovery. Coming globally via the Gemini API in @GoogleAIStudio and Gemini for business, and for Ultra subscribers in @antigravity, the @GeminiApp, and AI Mode in Search.
7
10
2
104
7,859
Gemini 4 Carbon is rolling out today. It’s built for sustained, long-horizon tasks, reasoning deeply through complex software engineering, cybersecurity, and scientific discovery. Coming globally via the Gemini API in @GoogleAIStudio and Gemini for business, and for Ultra subscribers in @antigravity, the @GeminiApp, and AI Mode in Search.
570
602
266
8,561
801,066
I joined Google DeepMind ~2 months ago! In SF & MTV for @aidotengineer! ​As a singleton in Singapore working on model launches & partners (team in EU/US), I learnt: ​1. Async work is autoregressive - continuous tokens in/out 2. In-person is latent diffusion 3. Trust is an efficient protocol 🧵
30
15
3
461
36,276
When building hard things: 1. Async feels autoregressive. You read a doc, reply to a chat, ship a PR. It’s fast and linear, but you only attend to what’s explicitly passed into your context window, even if it’s large. 2. In-person is latent diffusion. Overhearing a desk convo, wandering into a random talk, lunch chatter. It’s noisy, but somehow I get a much better sense of what to actually attend to. 3. Quote @benoitschilling - the tech will change, so focus on the people. Trust is an efficient protocol.
2
12
803
I’m curious to see how to bridge the remote gap. e.g. OpenClaw Multiplayer launched last week where teammates jump into the same live session. Radical transparency into what everyone’s working on, or chaos. Eitherway, excited to be here in person. If you’re at @aidotengineer this week, would love to say hi! 👋
9
511
Thanks for putting Gemini through the paces and honest feedback!
our team uses gemini quite a bit now whenever we have a tricky architectural problem, complex system design questions, or need to simplify something we ask for a gemini subagent it's pretty good here. i don't see this talked about much
2
1
113
5,936
Introducing Gemini 4 Carbon, it's a big jump over Argon which we released just 2 weeks ago. Gemini 4 Carbon is SOTA on cyber, long horizon tasks, ranks #1 on Artificial Analysis V5, significantly improves on usability in domains like coding, and is more aligned.
1,236
1,155
862
15,868
1,313,990
Excited to build with Gemini API CLI in Hermes, OpenClaw etc - one API for images, videos, tts, music makes it easy to manage
We launched the Gemini API CLI. A CLI for calling the Gemini API. Give it to the agent you already use (Claude Code, Codex, Antigravity, Cursor, OpenClaw, Hermes...). It's NOT a new coding agent you can use it too: 1. Generate images, video, speech, music. 2. Analyze PDFs, images, audio, video, YouTube URLs, 3. Transcribe recordings to md / json / srt, 4. Run Deep Research in the background. Stateless. No memory, no chat mode. Your agent owns the context and the plan, gemini-api does the API call. Build for Agents, it's self-describing. `--help`, `--dry-run`... Start here: aistudio.google.com/learn/in…
9
466