PrismML
@PrismML · United States · joined 28 Mar 2025
Centering AI research on efficiency. https://t.co/88MQHGCeFD
27 806Followers
31Following
337Posts total
2.7MViews on collected posts
Neueste Beiträge
Wir suchen neue Beiträge…
@PrismML 9x smaller with 98.2% retention is nuts
19.1K views · 458 likes
· 0 reposts · 4 replies
Open on X →
Building with Bonsai? We work with teams on domain-specific post-training and hardware optimization to meet memory, latency, and power constraints. Get in touch: contact@prismml.com.
74K views · 211 likes
· 11 reposts · 6 replies
Open on X →
Computer use is another strong test.
The model has to repeatedly interpret state, choose the next action, and stay coherent across a long sequence of steps, exactly where small capability gaps become visible.
Here is Bonsai 2 27B running a computer-use workflow locally on the h
111.6K views · 520 likes
· 23 reposts · 10 replies
Open on X →
2:30
Links:
Blog: https://t.co/0Xh3vu12Ex
Whitepaper: https://t.co/foPUfHCOV3
Models: https://t.co/M5Ymxjt9Wz
WebGPU Demo: https://t.co/Xq2QAoL0KT
GitHub: https://t.co/3BAp6Rg6az
Docs: https://t.co/MzBfoKhxh3
Discord: https://t.co/1MwtJroLJs
Join us: https://t.co/7TvnvrzO64
73.3K views · 509 likes
· 38 reposts · 17 replies
Open on X →
On an intelligence-density basis, Bonsai 2 27B is a clear outlier relative to both full-precision models and other low-bit alternatives.
Many low-bit models become deployable only with a meaningful quality tradeoff. Bonsai 2 pushes the frontier toward significantly higher https:
183.4K views · 553 likes
· 21 reposts · 12 replies
Open on X →
This is where higher retention matters most: fewer derailments across multi-step tasks, fewer silent failures as the state evolves, and better consistency from one decision to the next.
Here is Ternary Bonsai 2 27B running an agentic coding workflow with Cline on an NVIDIA https
131.9K views · 451 likes
· 26 reposts · 9 replies
Open on X →
1:19
♥️Huge thanks to @nvidia for the DGX Sparks for @openclaw engineering team.
We have been putting these to hard work, product announcements soon with our joint efforts to improve local model experience and empowering use-cases locally beyond what is currently possible on https:/
46.7K views · 333 likes
· 32 reposts · 24 replies
Open on X →
Try Ternary Bonsai 27B directly on Hugging Face , powered by @togethercompute
https://t.co/yB7uIaHXpF https://t.co/PEDLAI2C6u
18.8K views · 129 likes
· 15 reposts · 15 replies
Open on X →
0:23
Links:
Blog: https://t.co/GFFHOmzEf7
The Information: https://t.co/4xrfdMXoUH
CNBC: https://t.co/kwV3QAeH4v
Whitepaper: https://t.co/PwNgOpyuwV
Models: https://t.co/6sbYg2vcMl
WebGPU Demo: https://t.co/uNIoAZLFKt
API: https://t.co/slXyzECdnc
GitHub: https://t.co/3BAp6Rg6az
Docs:
33.6K views · 319 likes
· 33 reposts · 10 replies
Open on X →
The phone threshold is even harder than the storage number suggests.
A phone exposes only part of its memory to an application, and the model must share that budget with its KV cache, activations, runtime, and the rest of the product.
At 3.9 GB, 1-bit Bonsai 27B clears that htt
33.6K views · 219 likes
· 14 reposts · 9 replies
Open on X →
1:19
Here is Ternary Bonsai 27B running an end-to-end agentic workflow locally with Hermes on an NVIDIA GeForce RTX 5090 GPU.
The model reasons, calls tools, reads outputs, modifies files, and surfaces insights - all on consumer hardware, while all private files, intermediate states,
155.6K views · 547 likes
· 55 reposts · 11 replies
Open on X →
1:46
The footprint reduction does not come at the expense of the capabilities that matter.
Across 15 benchmarks spanning knowledge, reasoning, math, coding, instruction following, tool use, and vision, Ternary Bonsai 27B retains 95% of the full-precision model’s performance. The http
34.2K views · 272 likes
· 22 reposts · 5 replies
Open on X →
Why does this matter? Because modern AI workflows are no longer single prompts. They are sustained loops.
A capable agent may take hundreds of steps: reasoning, calling tools, reading outputs, updating its state, and iterating toward a result. When every step is remote, https://
27.4K views · 195 likes
· 13 reposts · 1 replies
Open on X →
Today, we’re announcing Bonsai 27B: the first 27B-class model to run on a phone.
Bonsai 27B is the new multimodal flagship of the Bonsai family. Based on Qwen3.6 27B, it brings a new capability tier to local AI: multi-step reasoning, structured tool use, long-context workflows,
1.5M views · 6K likes
· 921 reposts · 309 replies
Open on X →
Raw capability determines what a model can do. Intelligence density determines where it can do it.
Bonsai 27B moves the Pareto frontier left again: 27B-class capability in a footprint smaller than many full-precision 2B models.
By intelligence density, 1-bit Bonsai 27B delivers
40.8K views · 300 likes
· 14 reposts · 3 replies
Open on X →
Today we’re releasing 1-bit and Ternary Bonsai Image 4B.
A new family of image-generation models designed to run high-quality diffusion inference on local hardware: from laptops to phones. https://t.co/9qB5UbOogJ
211.4K views · 1.7K likes
· 235 reposts · 69 replies
Open on X →
Im Vergleich zu Konten gleicher Größe
17 Beiträge aus den letzten 90 Tagen, verglichen mit der Größenklasse 10K–100K Follower. wird weit gezeigt, aber nur wenige dieser Zuschauer reagieren.
Medianaufrufe40 809dieses Konto996Median für 10K–100K
Reichweite, %146.76%dieses Konto3.84%Median für 10K–100K
Interaktion, %0.77%dieses Konto1.51%Median für 10K–100K
| Kennzahl | Dieses Konto | Median für 10K–100K | Verhältnis |
|---|---|---|---|
| Medianaufrufe pro Beitrag | 40 809 | 996 | 41.0× |
| Reichweite (Aufrufe ÷ Follower) | 146.76% | 3.84% | 38.2× |
| Interaktionsrate | 0.77% | 1.51% | 0.51× |
Weitere Konten dieser Größe → Mit einem anderen Konto vergleichen → Wie diese Vergleichswerte entstehen →
Growth & engagement
How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.
Views per post
34.2K14 Jul
155.6K
33.6K
33.6K
5.3K
18.8K21 Jul
46.7K24 Jul
16.3K
131.9K17 Sep
183.4K
73.3K
111.6K
74K
19.1K
Last 14 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.
Engagement rate per post
0.89%14 Jul
0.40%
0.73%
1.09%
0.93%
0.86%21 Jul
0.84%24 Jul
0.37%
0.37%17 Sep
0.32%
0.77%
0.50%
0.31%
2.42%
Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.51%.
What the audience does
Likes56.1%12 753 in total
Reposts6.5%1 477 in total
Replies2.3%519 in total
Quotes3.1%703 in total
Bookmarks32.0%7 268 in total
Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.
Followers by day
19 Sep
Daily snapshots since 19 Sep 2026; the dashed line is the starting count.
Ähnliche Konten
Merlin 🎇 Griffiths ✓
@MerlinFDC4
55.6K followers
· 15.5K posts
big rob 🏴
@SimpleFacks
55.6K followers
· 13.2K posts
ar.io
@ar_io_network
55.6K followers
· 10.1K posts
Anthony Westgate ✓
@a_westgate
55.6K followers
· 15.7K posts
Kuromiya Lucien ✓
@kuromiyalucien
55.6K followers
· 31.6K posts
tkz כֶּשֶׁף ✓
@tuukzs
55.6K followers
· 122.1K posts
Gabs ✓
@cartasdogabs
55.6K followers
· 32.7K posts
Slick_ZM
@Slick_ZM
55.6K followers
· 76.4K posts