Gemini 4 Carbon early test results are in ahead of public release
Gemini 4 Carbon appears to be in early access outside Google, with testers citing stronger outputs than Argon but weaker UI work and much slower speeds. Leaks suggest launch may be close, though public claims remain unverified.
🗞 #google @testingcatalog
Gemini 4 Carbon appears to be in early access outside Google, with testers citing stronger outputs than Argon but weaker UI work and much slower speeds. Leaks suggest launch may be close, though public claims remain unverified.
🗞 #google @testingcatalog
TestingCatalog AI News
Gemini 4 Carbon early test results are in ahead of public release
Barely a week after Gemini 4 Argon went public, Google's next checkpoint, Carbon, is showing up on LMArena and with partners. Early testers say it's a clear step up in 3D work, with a few familiar weak spots.
MiniMax has published MiniMax M3.1 Flash weights on HuggingFace.
The model has 428B parameters total, 23B activated per token.
MiniMax M3.1 Flash, also known as "Space Bunny" stealth model from OpenRouter, was announced on October 19 and is now available on MiniMax platform and APIs.
👍7🔥2❤1
META 🔥: Muse AI agent is rolling out in Japan!
US, Canada, Japan..
We need it EU as well 👀
US, Canada, Japan..
We need it EU as well 👀
❤11🗿5👍4
ICYMI 👀: Perplexity released its pplx-decider-flash-v1.2 decision model as Mac application that serves decision API locally.
The model takes 5.8 GB of a space and requires 16 GB of memory.
100% True 🤖
The model takes 5.8 GB of a space and requires 16 GB of memory.
100% True 🤖
❤9
Media is too big
VIEW IN TELEGRAM
ByteDance released Seedance 2.5 4K, featuring richer visual detail.
> Seedance 2.5 4K offers richer color gradations and smoother transitions thanks to 10-bit color depth.
> Users can use Draft mode to experiment with low-cost 480p previews and then upscale them to 4K.
Announcement demo 👀
> Seedance 2.5 4K offers richer color gradations and smoother transitions thanks to 10-bit color depth.
> Users can use Draft mode to experiment with low-cost 480p previews and then upscale them to 4K.
Announcement demo 👀
❤4🔥2👀1
This media is not supported in your browser
VIEW IN TELEGRAM
SPACEXSI 🔥: Cursor now has Computer Use capabilities and can operate other applications on your laptop.
The feature is especially usefull for app development, as it can also run mobile simulators.
Computer Use is activated on demand or via @-mentioning "Computer Use" command.
Cursor is a Superapp now! 👀
The feature is especially usefull for app development, as it can also run mobile simulators.
Computer Use is activated on demand or via @-mentioning "Computer Use" command.
Cursor is a Superapp now! 👀
❤4
🚨 AI News | TestingCatalog
BREAKING 🔥: Kimi K3.1 Pro is now rolling out on Kimi AI web and mobile. > Kimi K3.1 Pro - Complex tasks and professional work, all-around flagship. Testing time! 👀
BREAKING 🔥: Kimi AI officially announced Kimi K3.1 Pro, a 2.8T total and 104B active params MoE open model.
On benchmarks, Kimi K3.1 Pro performs on par with leading AI models like Opus 5.5 and GPT-6 Astra, while being significantly cheaper.
Open weights are coming on November 4th as well!
On benchmarks, Kimi K3.1 Pro performs on par with leading AI models like Opus 5.5 and GPT-6 Astra, while being significantly cheaper.
Open weights are coming on November 4th as well!
❤6🔥4
🚨 AI News | TestingCatalog
BREAKING 🔥: Kimi AI officially announced Kimi K3.1 Pro, a 2.8T total and 104B active params MoE open model. On benchmarks, Kimi K3.1 Pro performs on par with leading AI models like Opus 5.5 and GPT-6 Astra, while being significantly cheaper. Open weights…
KIMI AI🔥: Kimi K3.1 Pro landed on the 8th spot on WebDev Arena, with Kimi AI moving to the 4th place among AI labs overall.
That a big move 👀
That a big move 👀
❤5👍3
OPENAI 🔥: Frontier AI training was already resumed on October 1st, according to the latest statement from OpenAI.
This is a very important part as the cost of aligned frontier AI development will go up proportionally to its intelligence. Eventually, this may impact token pricing as well.
Yet, even these measures might be not sufficient which would imply further cost increase.
“We leave in the era of underpriced compute.”
Additionally, a substantial amount of compute will be consistently allocated to red teaming.
This is a very important part as the cost of aligned frontier AI development will go up proportionally to its intelligence. Eventually, this may impact token pricing as well.
Yet, even these measures might be not sufficient which would imply further cost increase.
“We leave in the era of underpriced compute.”
❤7
Media is too big
VIEW IN TELEGRAM
Jack Dorsey announced a new full-duplex speech model, called Dessa.
Dessa was trained solely on voice conversations, with no text data or pretrained models.
Currently, only demo samples are available, but they far surpass anything else in emotional expression and natural feel.
Voice models will be everywhere 👀
Dessa was trained solely on voice conversations, with no text data or pretrained models.
Currently, only demo samples are available, but they far surpass anything else in emotional expression and natural feel.
Voice models will be everywhere 👀
❤6🔥2👍1