Senior software engineer Automation + applied AI (pharma) Contributor: MLX-Serve, TensorFold, ds4 Local inference · Real hardware · Builders
Joined September 2026
- Tweets870
- Following428
- Followers1.6K
- Likes5.8K
Sebastian Christiansen retweeted
ATTENTION DGX SPARK & MLX OWNERS:
The CUDA.fast | MLX.fast Challenge is HERE, and this round we're partnering with the a local community leader @ashxhart
The Challenge: Optimizing Qwen 3.8 27B on the TensorFold Engine. @tensorfold_dev
Point your agents at one of the challenge pages and start making optimizations and STAY on the leaderbaord.
Easy to hit the leaderboard...good luck staying number 1 though.
You up for the challenge? @MiaAI_lab @Tech2Wild
Sebastian Christiansen retweeted
Local AI Builders Friday # 2
The list is now 102. Every week this list sends me down a new rabbit hole.
I added 8 more amazing builders this week:
1) @divagr1925 built Qwenputer, a RISC-V CPU inside Qwen3.5-4B and uses the model’s own layers as RAM. It runs C and MicroPython in its forward passes with no external runtime. AI/ML engineer.
2) @mhelcig made the GSQ-RCO quants that put DeepSeek V4.1 Flash, a 763B open model, on a 12GB GPU, running in @coldniko's Strata engine. One user is already getting 40 tok/s on a single 5090.
3) @machinestein open sourced Locals, an Apple Silicon engine for local LLMs. With LFM3 added this week it hit 620 tok/s on a 16GB M2 running 8 agents in parallel. That’s 3x mlx-lm.
4) @adrgrondin makes Locally AI, which Apple just featured for the new iPad mini. This week he helped bring oMLX, an MLX engine built for long context, into LM Studio and Bionic.
5) @AlicanKiraz0 spent a week with ASUS’s GB300-powered ET900N G3 and wrote the review in 7 languages. In his ant colony coding benchmark, DeepSeek V4.1 Flash beat Kimi K3.1 Pro for 17 cents.
6) @sfxnz got a local model on 2 x DGX Sparks to build an Apple TV-style launcher for his Samsung TV. Keeps building with Local AI.
7) @CerebralCoding_ builds Latch, a scheduler so agents stop fighting over the same GPUs and contributes to MLX-Serve and TensorFold to get models running on Apple Silicon.
8) @lasse_ai tested clock speeds on two DGX Sparks. At 2200 MHz they keep almost full speed and at 1900 MHz they run cooler and quieter.
Follow them and follow the full list here: nitter.gq/i/lists/21089799789857…
Who should be in #3?
Local AI Builders Friday #1 🦙
On Oct 10, I highlighted 24 builders from my Local AI Builders list. The list is now 94!
Starting this week, as part of this series, every Friday I'll shout out new Local AI people and the ones I've missed.
So many OGs and so many cool people are building at the intersection of Local AI, MLX, open source AI, open-weight models and quantization.
Worth a follow:
- @hnshah
- @gerdelgado
- @ClementDelangue
- @julien_c
- @TeksEdge
- @bleysg
- @Teknium
- @haider1
- @HessianFree
- @SteveSTechMan
- @u1tra_instinct
- @mervenoyann
- @mchiang0610
- @wightmanr
- @alexocheema
- @thebasedcapital
- @VikParuchuri
- @_philschmid
- @turboderp_
- @natolambert
- @shshnkp
- @macintogdev
- @ramin_m_h
- @0xSigil
- @theemozilla
- @digitalix
- @goodhunt
- @Technovangelist
- @Hikari_07_jp
- @code_hiyouga
- @tcarambat
- @_LuoFuli
- @zhuohan123
- @ActuallyIsaak
- @gladkos
- @philipkiely
- @ornith_
- @xenovacom
- @MaziyarPanahi
- @SahinLale
- @lgtm_hbu
- @sudoingX
- @jartine
- @woosuk_k
- @jeremyphoward
- @stevibe
- @StasBekman
- @FrostForger
- @DanKornas
- @melvindvivas
Pin it and you get a feed that's only Local AI.
Full list: nitter.gq/i/lists/21089799789857…
Who else should I add?
Sebastian Christiansen retweeted
spark-ds41f-1030 is out. DeepSeek V4.1 Flash on DGX Sparks. 1M context. Native weights.
Two models at once: stock and abliterated, one server. Nobody touched the weights.
Tuned for multi-session interactivity: short requests jump the queue ahead of long prefills.
Less memory wastage, more memory headroom.
Oh. And it's faster again.
Huge thanks to Jordi, who found our three-Spark tuning and qualified it on his own triangles before we did.
github.com/christopherowen/s…
Sebastian Christiansen retweeted
1) local is about freedom.
2) 5% utilization is not good. thank you for making the case for darkbloom on national podcast, sama
sam: local inference is a terrible allocation of scarce resources
“a lot of people want local inference, and somehow they assume that the problem is solved now because we have open-source models, we're going to be able to run them on our computers, and so we're not going to need all the stuff in the data centre
but from a global allocation perspective, actually it's quite terrible to have everybody, like, have a bunch of the world's limited supply of DRAM sitting in their own computer and being used five percent of the time
you want maximum utilisation in the cloud if you're trying to deal with this very supply-constrained environment
somehow many people's intuition is the opposite of that”
Sebastian Christiansen retweeted
I’ve got a really good friend, like top tier dude I promise you and coming from me it means something, who’s going through a tough time through no fault of his own
He’s a great full stack dev, honestly I’d hire him if I had any work to throw at him
If you have any work you need help with please contact me and I will put you in relation
You can hold me accountable if anything happens and I will reimburse you
That’s how much I trust this guy
Sebastian Christiansen retweeted
Most hardworking people in a Microsoft office
This quoted post is unavailable.
Sebastian Christiansen retweeted
Hi 👋 Neywa Labs here.
Our mission is to make AI available to everyone in the world.
Building something big? We're collaborating with design partners, come talk to us.
Newyalabs.ai
Today we're coming out of stealth, introducing Neywa Labs 🚀
We built the infrastructure to build and deploy AI that runs on the devices people already own. Private, offline, unlimited.
One mission: make AI available to everyone in the world.
neywalabs.ai/
Sebastian Christiansen retweeted
EBITDA = Earnings Before I Trick Da Auditor
Anthropic has filed an S-1 registration statement with the SEC related to its proposed IPO.
Read more: anthropic.com/news/s1-sec
Legal disclaimer: A registration statement relating to these securities has been filed with the SEC but has not yet become effective. These securities may not be sold, nor may offers to buy be accepted, prior to the time the registration statement becomes effective. This post does not constitute an offer to sell or the solicitation of an offer to buy these securities, nor shall there be any sale of these securities in any state or jurisdiction in which such offer, solicitation or sale would be unlawful prior to registration or qualification under the securities laws of any such state or jurisdiction.
Sebastian Christiansen retweeted
If your company refers to engineers as the IT department, you have nothing to worry about. You were already breached last week.
Sebastian Christiansen retweeted
STOP STOP STOP STOP!
You need to try my BlackFold TP2 DeepSeek V4.1 1017 build. Uncensored on TensorFold
YOU actually need to try it right NOW.
Look at what happens when you push it to xhigh reasoning:
🔥 90/100 on tool eval bench (173/192)
⚡ ~70 tokens/sec avg coding speed locally
Huge shoutout to @ivanfioravanti for putting it through its paces and showing what this beast can do outside my own test environment.
If you want absolute raw performance with near-zero refusals running locally on your hardware, stop scrolling.
Model and setup links below. Run it, push it, and show me what you build 👇
DeepSeek-v4.1-Flash-1017-heretic-tf correction! BlackFold test below was with:
--reasoning-effort none.
With xhigh it got 90/100. 173/192 points!
Still Adequate as rating, but mainly for safety, that clearly is irrelevant for this model and Structured Output where it seems Main cause: TensorFold rejects response_format engine rejects both json_schema and json_object. @Blackwellboy is looking into this.
I'm using it for coding right now and it keeps ~70 t/s as average! 🔥
