Thinking, posting AI n stuff.
Joined January 2022
- Tweets 532
- Following 45
- Followers 501
- Likes 961
201 Photos and videos
Pinned Tweet
13 Nov 2025
π¨ GEMINI 3.0 PRO - macOS
This model is truly insane... π€―
Gemini 3.0 and Nano Banana 2 dropping at the same time would be so perfect... imagine the possibilities of them both without the restrictions/limits π«£
50
93
1,706
269,637
Jun 9
π¨ Claude Fable 5 OUT NOW!
Anthropic has released their Mythos class model Fable 5!
$10/1M input - $50/1M output
The ball is in OpenAI's court now for 5.6 π
Introducing Claude Fable 5: a Mythos-class model that weβve made safe for general use.
Its capabilities exceed those of any model weβve ever made generally available.
1
4
1,119
May 29
π¨ Codex Computer use on Windows!
Finally, huge for lots of users that don't have a Mac - hopefully we see more Windows support in the future π₯
Windows users, this oneβs for you.
Computer use now works on Windows, so Codex can take action on your Windows computer.
And with Windows support for Codex in the ChatGPT mobile app, you can start, review, and steer tasks on the go while work continues on your Windows machine.
An early experience, but weβre working on more ways to keep your work moving, wherever you are.
7
1,439
May 28
π¨ Opus 4.8 OUT NOW!
Same price as Opus 4.7 but smarter, more honest and better at long running tasks! π
Introducing Claude Opus 4.8: it builds on Opus 4.7 with sharper judgment, more honesty about its own progress, and the ability to work independently for longer than its predecessors.
Available today at the same price.
ALT Benchmark table showing how Claude Opus 4.8 compares to its predecessor and to other models on tests of coding, agentic skills, reasoning, and practical knowledge work tasks.
1
5
1,610
May 19
π¨ Gemini Spark out now
Personal 24/7 AI agent using 3.5 Flash, a definite step toward AGI imo
$100 AI plan out now too and Ultra has been reduced to $200! π₯
Introducing Gemini Spark β¨
Itβs your 24/7 personal AI agent that helps you navigate your digital life, taking action on your behalf, and under your direction.
π§ It runs on Gemini 3.5 and is built on @Antigravity, so it can perform long-running tasks easily in the background.
β±οΈ And because it runs on dedicated virtual machines on Google Cloud, you donβt even need to keep your laptop open.
π§° Spark will integrate seamlessly with Google tools, and soon with third parties through MCP.
#GoogleIO
7
826
May 19
π¨ Gemini 3.5 Flash Out Now
400 Tokens per second with subagents in Antigravity
3.5 Pro out next month π
Meet Gemini 3.5 Flash β our strongest agentic and coding model yet.
It delivers frontier-level performance at 4x the speed of comparable frontier models β often at less than half the cost.
Generally available, starting today. π§΅
#GoogleIO
ALT Text that says Gemini 3.5 Flash set against a deep blue background.
3
225
May 3
Interactive Ferrari Simulation Website π
GPT-5.5 Images V2 inside Codex can do some great stuff, here I had Images V2 generate the entire design of the site, cars and panoramic interiors then used Apple SHARP to turn the cars interiors into guassian splats to get that 3D feel π₯
1
13
3,511
Apr 27
π¨ Opus removed from Claude Code
Anthropic now require $20 Pro plan users to enable purchase extra usage to use Opus models on Claude Code...
It's like they want people to move to Codex π€
6
11
1,521
Apr 24
π¨ DEEPSEEK V4 OUT NOW
DeepSeek have come out of the blue with 2 open source models, V4 Pro (1.6T, A49B) and Flash (284B, A13B) with Pro Max variant for maximum reasoning!
huggingface.co/deepseek-ai/Dβ¦
1
7
901
Apr 23
π¨ GPT-5.5 OUT NOW
Full benchmarks - but don't let them fool you, the model genuinely has a lot more understanding and feels like a proper leap when in real use
Time to build π₯
Introducing GPT-5.5
A new class of intelligence for real work and powering agents, built to understand complex goals, use tools, check its work, and carry more tasks through to completion. It marks a new way of getting computer work done.
Now available in ChatGPT and Codex.
1
11
1,176
Apr 22
π¨ Agents have been released in ChatGPT!
OpenAI's vision for AGI becomes clearer with every update, the combination of intelligence and its capabilities are getting stronger rapidly π
Introducing workspace agents in ChatGPTβshared agents that can handle complex tasks and long-running workflows across tools and teams.
7
1,580
Apr 22
GPT-5.5 confirmed for Thursday π
The combination of 5.5 and Images v2 should allow for some really good stuff especially in design π₯
2
8
1,292
Apr 21
π¨ GPT-5.5 Pro aka Spud - Transformer Flight Sim
WOW!! π€―
This is the best result I've ever seen for a flight sim, homing rockets, fly into space, transform into a robot, flares and more...
OpenAI have COOKED! π§βπ³
2
13
1,487
Apr 20
OpenAI are lying about the $100 Pro plan in ChatGPT
They claim unlimited usage of all GPT-5 models as long as you adhere to the Terms of Use... but the Pro model only has 50 queries a week on the 5x plan π€
This just comes across as deceptive marketing, hopefully we get some clear communication regarding this.
1
14
2,705
Apr 20
π¨ GPT-Image 2 is rolling out now on ChatGPT!
Here's an easy prompt to test if you have it or not as V2 is the ONLY model that can do this:
"an ingame screenshot of rust"
Left image = GPT-Image 1.5 (Rust/CoD/Tarkov mix)
Right image = GPT-Image 2 (Literally Rust)
10
2
31
13,176
Apr 20
Feels like OpenAI trained Images V2 on literally everything...
It has great understanding, can do stuff no other model can and is definitely SoTA for generation π₯
Should be a good week π
12
1,088
Apr 19
π¨ GPT-5.5 Pro - Sam Altman/OpenAI website
The model is really good at understanding images/videos, this site was made with a video reference and did it really well
Still behind Gemini and Claude when it comes to designing without a reference though unfortunately
3
2
21
3,041
Apr 19
π¨ GPT-5.5 PRO - 3D Voxel Pokemon Duel
HOLY $#!T π€―
One of the best generations I've ever seen from a model, it just works and looks great... OpenAI have truly cooked π
3
2
27
2,967
Apr 19
π¨ GPT-5.5 PRO - Geometry Dash
GPT-5.5 Pro aka Spud is the ONLY model that can do this correctly from what I've tested π
Each level is actually completable, has good logic, fluidity and even added the different modes like ball and ship with it's own music for each level π€―
2
3
34
3,647