Joined April 2016
- Tweets 6,056
- Following 253
- Followers 409
- Likes 6,147
830 Photos and videos
Pinned Tweet
4 Oct 2025
POV:
This is Bredda, Full-Stack ADHD Developer and Engineer 👋
He is the un-hinged side of development...
From half-baked projects to the ones that actually ship (miracles do happen)
Follow and you might see one!
bredda.dev/
9
1,976
23h
Hey guys!
A friend of mine create this amazing site for checking out the average TPS for each AI model from most provider.
This can finally allow you to see where you can get the best TPS for your favourite model!
Any feedback would be appreciated
aispeedometer.live/
1
1
2
16
Jun 13
YEAAAAAH BOI
MiniMax M3 3x usage on @opencode really great model and now its value is insane!
3
44
Jun 13
I was SOO close to buying the sub... This would of been absolutely hilarious if I subbed and fable got removed...
Jun 13
As a result of a US government directive, we are suspending access to Claude Fable 5 for all users. You can continue to use all other Claude models.
Here’s what this means for you:
Across Claude products, new sessions will run on your selected default model or Opus 4.8, and existing Fable 5 sessions will end with an error.
On the Claude Platform, requests to Fable 5 will also return an error. Please update your integrations to other Claude models.
We know this is a disruption to your workflows; we appreciate your patience and support.
1
1
19
Jun 12
Finally the open weights are released for a really good model that I love!
Jun 12
MiniMax M3, Open-Weight, Now On Hugging Face , with only ~428B parameters and ~23B activated parameters
Weights:
huggingface.co/MiniMaxAI/Min…
MiniMax Sparse Attention:
huggingface.co/papers/2606.1…
11
Jun 12
A respectable 0.1 increment improvement on a respectable model...
Glad to see company taking time to release minor updates to improve quality of life rather than chasing benchmarks...
This probably would have more affect than having a higher bench mark score
Jun 12
🌘 Kimi-K2.7-Code, our latest coding model, is now released and open-sourced!
🔷 Improved coding & agent performance over K2.6: 21.8% on Kimi Code Bench v2, 11.0% on Program Bench, and 31.5% on MLS Bench Lite.
🔷 Reasoning efficiency: Less overthinking, with 30% lower reasoning-token usage compared to K2.6.
🔷 Long-horizon coding: Improved instruction following, higher end-to-end coding task success rates.
⚡️ 6x High-Speed Mode coming soon!
🔌 Available today via Kimi API and Kimi Code.
🔗 Kimi Code: kimi.com/code
🔗 API: platform.moonshot.ai
2
8
Jun 11
It seems that we are being slowly introduced a dislike button on comments
21
Jun 9
it finished before I was able to finish this post!
Jun 9
Boys...
MiMo V2.5 Pro Ultraspeed is doing files per seconds...
not tokens per second.. 🤯
10
Jun 9
Boys...
MiMo V2.5 Pro Ultraspeed is doing files per seconds...
not tokens per second.. 🤯
1
86
Jun 9
How generational debt is generated...
25
Jun 8
New MiMo UltraSpeed comparison.. you literally cant even think this fast!
60
Jun 8
MiMo is now at 1000 Tokens per second!!
This is insane speed! the amount of work this can do it UNBELIEVABLE at this speed...
Jun 8
🚀 1,000 TOKENS/S ON A 1T MODEL! 🚀
We are thrilled to release Xiaomi MiMo-V2.5-Pro-UltraSpeed in collaboration with @TileRT_AI , breaking the 1,000 tokens/s output speed on a 1 Trillion parameter model for the FIRST TIME!
Not wafer-scale integration like Cerebras. Not pure on-chip SRAM chips like Groq. We achieve 1,000 tps on a 1T MoE model using just a SINGLE, STANDARD 8-GPGPU NODE.
Read the full technical deep dive：mimo.xiaomi.com/blog/mimo-ti…
Want to experience the future of real-time AI?
👉 Apply for UltraSpeed now: platform.xiaomimimo.com/ultr…
⏳ Limited-Time Access: Application-based · Jun 8 – Jun 23 (PDT)
💬 Chat Experience: Completely FREE for a limited time — try the blazing-fast web chat now.
⚡ UltraSpeed API: Just 3x the price for a ~10x boost in output experience.
🤝 Enterprise & Large-Scale Needs: business-mimo@xiaomi.com
68
Jun 8
keep in mind boys..
despite spending over a million dollars in tokens a month.. they still lost to @NousResearch...
dont fall to the tokenmaxxing trap..
Jun 7
Here’s your monthly reminder that you shouldn’t be prompting coding agents anymore.
You should be designing loops that prompt your agents.
1
43
Jun 8
by the end of the year.. I suspect we are going to see a $500 / $1000 plans coming out...
Jun 8
Why do I get the feeling that $100 AI plans are becoming more and more like previews..
13
Jun 8
I was gonna make this idea but stopped because I dont know how many people would need it..
So I will share it and we will see..
Imagine Kickstarts Venture captial comes together!
creating an open VC marketplace where users can crowd fund projects and gain shares from that.
companies would follow a guided structure with cap tables, phases, milestones, etc.. to follow to unlock the funds.
while users who angle invest can invest as much as they want and be able to sell their shares at any time.
in the golden age of anyone being able to make a startup. funding becomes a heavy limiting factor.
this way you people can raise funds a lot easier and gain exposure.
I would like to make something like this but from my research, something similar to this already exists.. also I dont have the connections and the network to be able to gain the trust and users to use this..
hope someone would do this..
FYI, openvcx.com is an available domain!
39
Jun 7
testing out the Hermes Desktop app on Windows with a random task using Nemotron 3 Ultra
the Desktop app is REALLY good.. I am loving the design and how simple it is... if this can become a mobile app as well.. this might be an ultra cook..
this is giving me codex vibes but with much more flexibility and model selection.
Not to forget boys.. this can connect to your VPS instances too so it can also manage them!
would love to see an option where I can have multiple VPS managed in this so each chat has its own hermes agent in a VPS for real ULTRA work..
1
1
86
Jun 7
@Teknium would love if you would be able to make this work...
where users can have multiple VPS connected and have grouped chats based on the VPSs that users have running..
so I can control all of my different Hermes through this app and have it where I can do many different things at once.
2
36
Jun 4
I am seeing a LOT more reply's than before...
something different with this algo..
11