馃毃Google just revealed Genie 4
Google's latest world model generates interactive environments in real time, letting you explore virtual worlds and see them respond to your actions.
It's currently waitlist-only, with wider access coming later.
Google's latest world model generates interactive environments in real time, letting you explore virtual worlds and see them respond to your actions.
It's currently waitlist-only, with wider access coming later.
0
0
0
572
馃毃 Is GPT-6.1 Terra draining usage faster than Astra?
I'm on the $100 Pro plan, and Terra High has already used 18% of my quota in just 2 hours.
What's strange is that Terra was expected to be cheaper and more efficient.
Is Terra actually more expensive to run in practice, or does it depend on the workload and reasoning effort?
Anyone else noticing faster usage drain with Terra compared to Astra?
I'm on the $100 Pro plan, and Terra High has already used 18% of my quota in just 2 hours.
What's strange is that Terra was expected to be cheaper and more efficient.
Is Terra actually more expensive to run in practice, or does it depend on the workload and reasoning effort?
Anyone else noticing faster usage drain with Terra compared to Astra?
14
2
47
5,452
馃毃 New OpenAI model spotted on Arena 馃憖
"Fe-pichu" uses OpenAI's tokenizer
appears to be under testing right now.
A new Qwen model has also surfaced on Arena.
"Fe-pichu" uses OpenAI's tokenizer
appears to be under testing right now.
A new Qwen model has also surfaced on Arena.
0
0
16
1,419
Did Codex just NERF Astra's usage limits AGAIN????
Something feels off with Astra XHigh lately. On Pro 5脳, I used to get roughly five full workdays out of my weekly quota. Today, just 2.5 hours into coding, 42% is already gone.
The bigger issue seems to be speed. Astra now spends ages executing commands and reasoning through tasks that used to be straightforward.
Is the quota being drained faster, or has Astra become noticeably less efficient?
Curious if other Pro users are seeing the same thing.
Something feels off with Astra XHigh lately. On Pro 5脳, I used to get roughly five full workdays out of my weekly quota. Today, just 2.5 hours into coding, 42% is already gone.
The bigger issue seems to be speed. Astra now spends ages executing commands and reasoning through tasks that used to be straightforward.
Is the quota being drained faster, or has Astra become noticeably less efficient?
Curious if other Pro users are seeing the same thing.
Day 22: GPT-6.1 Terra shows progress in writing quality and comprehension.
Is Terra finally getting better, or is the difference barely noticeable?
7
2
21
2,405
@CodexResets1 The real concern is whether the usage limits changed or the model is becoming less efficient. Pro users deserve more transparency on both.
1
0
0
99
TypeScript catches a lot of mistakes before your code runs.
Python lets you move incredibly fast.
Rust gives you stronger guarantees.
If you were starting a new backend today, which trade-off would you choose?
Python lets you move incredibly fast.
Rust gives you stronger guarantees.
If you were starting a new backend today, which trade-off would you choose?
40
2
48
993
Day 22: GPT-6.1 Terra gets another update, with better writing and comprehension.
Noticeable upgrade or still nothing special?
Noticeable upgrade or still nothing special?
Day 22/ Terra is back!
Our first model that is significantly better at writing, personality and nuanced understanding of requests.
Packs a punch across the board, very good at computer use, and gives 2X more usage than 6.1 Sol across API and subscriptions.
3
1
13
2,553
GPT-6.1 Terra is here, but will it earn a permanent spot in your workflow?
Are you planning to test it seriously, or sticking with your current model?
Are you planning to test it seriously, or sticking with your current model?
GPT-6.1 Terra: delivering fast and capable computer use at a lower price.
GPT-6.1 Terra understands what鈥檚 on screen, takes action across apps, writes naturally, and is a more capable model for your whole workday with the efficiency to take on more of it.
2
0
21
2,721
馃毃 GPT-6.1 Terra scores just 48 on the AA Intelligence Index 馃拃
OpenAI might be in trouble with this one 馃槶
OpenAI might be in trouble with this one 馃槶
42
6
443
92,249
馃毃 GPT-6.1 Terra benchmark results are out
Not gonna lie, these scores aren't looking too impressive so far 馃槶
Not gonna lie, these scores aren't looking too impressive so far 馃槶
2
1
44
8,276
Imagine building a $ 200M+ company...
...and then taking a software engineering job.
Why are founders doing this? 馃憖
...and then taking a software engineering job.
Why are founders doing this? 馃憖
201
26
1,462
777,415
if we鈥檙e being honest here, what does openai have going for them at this point? their products are poor, their infrastructure sucks and continues to get worse, and their models are lagging. 6.1 astra is going to have to be a big jump
192
34
1,757
160,559
@synthwavedd If they release a mediocre gpt 6.1 Terra at the end of these 28 days, we now that they are cooked.
0
0
5
556
馃毃 Fable 5.5 incoming?
If it delivers on the hype, we could be entering a new era of AI, where anyone can build businesses, create wealth, and solve problems once thought impossible.
Now, what is OpenAI preparing in response?
If it delivers on the hype, we could be entering a new era of AI, where anyone can build businesses, create wealth, and solve problems once thought impossible.
Now, what is OpenAI preparing in response?
馃毃 GPT-6.1 Terra: Leak could drop this week
Spotted in Microsoft evals and GitHub鈥檚 Copilot review workflow, Terra could match or beat Sol at a lower cost with better efficiency.
If it鈥檚 also faster than Sol, the price-performance could be insane. Codex bank resets too?
1
0
15
2,605
馃毃 GPT-6.1 Terra might be OpenAI鈥檚 next big surprise
- Cheaper than Sol, priced at $2/$10 per million tokens
- Rumored to outperform GPT-6.1 Sol
- 18 Microsoft eval setups moved from GPT-5.6 Terra to gpt-6.1-terra
- Could arrive within days
A model positioned below Sol beating it for less money would completely challenge OpenAI鈥檚 usual naming hierarchy.
If real-world tests confirm this, Terra could be far more capable than its name implies.
OpenAI might be cooking something serious.
- Cheaper than Sol, priced at $2/$10 per million tokens
- Rumored to outperform GPT-6.1 Sol
- 18 Microsoft eval setups moved from GPT-5.6 Terra to gpt-6.1-terra
- Could arrive within days
A model positioned below Sol beating it for less money would completely challenge OpenAI鈥檚 usual naming hierarchy.
If real-world tests confirm this, Terra could be far more capable than its name implies.
OpenAI might be cooking something serious.
18
5
217
28,882