Deepseek Flash is very hard to beat on easy tasks
- ship updates every month
- 10x cheaper than sonnet
- handles multi modal outputs
- ultra fast options will soon be available
Their performance and price combination will almost become impossible to match
- ship updates every month
- 10x cheaper than sonnet
- handles multi modal outputs
- ultra fast options will soon be available
Their performance and price combination will almost become impossible to match
96
10
268
12,854
Vibe Code Your Own LLM
You won't need a P.hD to create your own LLM and custom tune it to your use case anymore
You can simply go from a text prompt to a custom-trained LLM in an hour.
Simply ask the Abacus AI SuperComputer and use a GPU to train your custom weights in hours! Make your very own LLM
Coming next week
You won't need a P.hD to create your own LLM and custom tune it to your use case anymore
You can simply go from a text prompt to a custom-trained LLM in an hour.
Simply ask the Abacus AI SuperComputer and use a GPU to train your custom weights in hours! Make your very own LLM
Coming next week
94
20
360
22,053
🚨 Smaug-Nemotron is the best US open-source model in the agentic category
- can run locally on your desktop!
- tops agentic coding in US open-source
- number #1 on cost/task
- builds on Nvidia's latest model
We are excited to push the boundaries of US open-source. Weights available on HF
- can run locally on your desktop!
- tops agentic coding in US open-source
- number #1 on cost/task
- builds on Nvidia's latest model
We are excited to push the boundaries of US open-source. Weights available on HF
82
20
208
14,859
🚨 Kimi K3.1 Scores Below DeepSeek Flash
- seems to be bench-maxxed on public benches
- competitive on price per task
- it's not Fable or Astra class
- Deepseek is still king of open-source
This is on LiveBench
- seems to be bench-maxxed on public benches
- competitive on price per task
- it's not Fable or Astra class
- Deepseek is still king of open-source
This is on LiveBench
89
51
806
55,689
Looking at public benchmarks is a total waste of time
No lab worth its salt will release a model without checking and optimizing for benchmarks
Literally every model is optimized for benchmarks and 3D renders that can go viral on X
No lab worth its salt will release a model without checking and optimizing for benchmarks
Literally every model is optimized for benchmarks and 3D renders that can go viral on X
48
12
145
12,755
Kimi 3.1 is cheaper, but it's still not very competitive
It's priced at $2 / $10 per 1M input/output
Cheaper than K3 but not by 5x. DeepSeek still wins the open-source model race
It's priced at $2 / $10 per 1M input/output
Cheaper than K3 but not by 5x. DeepSeek still wins the open-source model race
48
12
432
20,105
The cost of intelligence is going to go down by 10x
Kimi K3.1 is going to kick it off by offering frontier level models at a fraction of the price
Chinese open source labs like DeepSeek will continue to get more performant while maintaining prices
Kimi K3.1 is going to kick it off by offering frontier level models at a fraction of the price
Chinese open source labs like DeepSeek will continue to get more performant while maintaining prices
60
17
399
13,187
Slightly bearish on SuperIntelligence until we crack robotics
Without robotics we are in the endless treadmill of low value digital loops
AI writing docs for other AI to read, writing code for other AI to use and so on 🤷♀️
With robotics, we begin to generate exponential value rapidly
Without robotics we are in the endless treadmill of low value digital loops
AI writing docs for other AI to read, writing code for other AI to use and so on 🤷♀️
With robotics, we begin to generate exponential value rapidly
90
10
201
12,715
Open source is so back 🚀
Kimi K3.1 drops shortly and will be great at 3D renders
LOL! People think it may rival Fable 5.5
I suspect it beats Argon and will place just behind OpenAI and Astra
Kimi K3.1 drops shortly and will be great at 3D renders
LOL! People think it may rival Fable 5.5
I suspect it beats Argon and will place just behind OpenAI and Astra
58
15
441
15,025
🚨 Announcing Smaug-Nemotron-3.5 - The Best US Open-Source LLM That Can Run On Your Desktop
We are super excited to push the boundaries of US open-source.
Smaug-Nemotron improves on Nvidia's latest model that literally runs on a single machine; all you need is one DGX Spark,
The Smaug variant is optimized for agentic coding, improves Deep-SWE, and enables local agents.
It's the best US agentic open-source model that can run on a single machine. Weights are now available on HuggingFace
We are super excited to push the boundaries of US open-source.
Smaug-Nemotron improves on Nvidia's latest model that literally runs on a single machine; all you need is one DGX Spark,
The Smaug variant is optimized for agentic coding, improves Deep-SWE, and enables local agents.
It's the best US agentic open-source model that can run on a single machine. Weights are now available on HuggingFace
83
22
374
3,401,784
Weird, OpenAI says they have internal models that are way better than their public models
If so, they should release some of those models and see if they can surpass Anthropic
Instead, they are postponing their IPO and securing $30B in "bridge financing" to tide them over 🤷♀️
If so, they should release some of those models and see if they can surpass Anthropic
Instead, they are postponing their IPO and securing $30B in "bridge financing" to tide them over 🤷♀️
48
3
217
14,219
TBH, the best way to stop the AI panic is for one of these frontier models to drop
- Fable 5.5
- Gemini Carbon
- Astra 6.1
Anthropic will probably be the first one with Fable 5.5 and will steady the stock market 😉
- Fable 5.5
- Gemini Carbon
- Astra 6.1
Anthropic will probably be the first one with Fable 5.5 and will steady the stock market 😉
32
7
240
12,765
OpenAI appears to be doing some real heavy lifting
They are running swarms of agents that are trying to hack into every system
Once the AI finds an exploit, you can patch it
The is so much better than random fear mongering and restricting access to models
They are running swarms of agents that are trying to hack into every system
Once the AI finds an exploit, you can patch it
The is so much better than random fear mongering and restricting access to models
94
13
404
18,093
🚨 RIP PowerPoint - Use AI For Animated Interactive Stories
- prompt to animated stories
- live graphs to showcase data
- uses Argon and Opus 5.5
- give up on boring slides
- upgraded to animated stories
- prompt to animated stories
- live graphs to showcase data
- uses Argon and Opus 5.5
- give up on boring slides
- upgraded to animated stories
42
99
1,896
3,467,147
After cracking all the math problems, Frontier Labs is running an unprecedented operation in the history of computing
They have deployed teams of millions of agents powered by their top models to find vulnerabilities
So yes, the good news is that they found some, and now everyone is scrambling to fix it!
They have deployed teams of millions of agents powered by their top models to find vulnerabilities
So yes, the good news is that they found some, and now everyone is scrambling to fix it!
43
14
232
15,823
The internet is getting swallowed by AI bots and agents
Terrified that I am responding to bots all day
Hello if you are human! Please say hi 👋
Bots please don’t reply
Terrified that I am responding to bots all day
Hello if you are human! Please say hi 👋
Bots please don’t reply
257
6
356
22,046
Anthropic’s philosophy is scary!
They want to be at the frontier so they can dictate how AI should be regulated
But absolute power inevitably corrupts absolutely
This could reach to extreme dystopia and a duopoly
The only good news is the other labs are investing billions and are almost at the frontier too 😓
They want to be at the frontier so they can dictate how AI should be regulated
But absolute power inevitably corrupts absolutely
This could reach to extreme dystopia and a duopoly
The only good news is the other labs are investing billions and are almost at the frontier too 😓
49
11
213
12,802
Argon is extremely good as a video director
Unlike Opus or Astra, Gemini's models are trained on the YouTube corpus
Argon can direct video models like SeeDance and audio models from ElevenLabs to create viral videos
Experimenting with it on our AI studio
Unlike Opus or Astra, Gemini's models are trained on the YouTube corpus
Argon can direct video models like SeeDance and audio models from ElevenLabs to create viral videos
Experimenting with it on our AI studio
57
30
709
32,120