Sujee Maniyam @sujee_dev

@deepseek_ai V4.1 Flash is now available on @nebiustf Here’s a quick demo in the Playground, including a speed test and a vision example. Links and credits in the reply 👇

2 1 4 102 0

View Details

Sujee Maniyam @sujee_dev

7 days ago

@DhruvDiddi @robertherobot @circuitlaunch haha, not sure who is more impressive.. human @DhruvDiddi doing pushups :💪 or robot 🤖 with the moves!

1 0 2 42 0

View Details

Arindam Majumder 𝕏 @Arindam_1729

a week ago

I was testing DeepSeek V4.1 Flash on @nebiustf , and honestly, the model is pretty damn good. So I built a tiny real-time webcam app that takes multiple frames from the camera and uses the model to summarize what it sees. The whole thing was surprisingly smooth, and the results were quite impressive for such a small setup. Here’s a quick demo of it:

Nebius Token Factory @nebiustf

a week ago

DeepSeek-V4.1-Flash is now live on Nebius Token Factory The model comes with native image understanding plus an architecture built for faster inference and higher throughput on long, input-heavy workflows. It combines native vision and a 1M token context with a 552B MoE that

5 6 52 17K 2

2 4 13 2K 2

View Details

Sujee Maniyam @sujee_dev

a week ago

@Yuchenj_UW My fav setup now @opencode + glm 5.3 on @nebiustf

0 0 0 17 0

View Details

Sujee Maniyam @sujee_dev

a week ago

@hunterguo101 💯 @hunterguo101

0 0 0 12 0

View Details

Sujee Maniyam @sujee_dev

a week ago

Want to try open models for coding? Keep your coding agent - Claude Code, Codex, OpenCode, Cline - and just swap the model underneath. In this short talk: * open vs closed models * why the coding harness matters * what coding tasks cost * quick setup demo Watch: youtube.com/watch?v=gjNcSo…

3 0 7 264 1

View Details

Sujee Maniyam @sujee_dev

a week ago

Setup + instructions: github.com/sujee/workshop… Want to try open models? Get credits: dev.nebius.com/builders?utm_s…

0 0 1 30 0

View Details

Sujee Maniyam @sujee_dev

a week ago

Try these models with free credits : dev.nebius.com/builders?utm_s…

0 0 0 14 0

View Details

Sujee Maniyam @sujee_dev

2 weeks ago

2 really good open models - GLM-5.3 and DeepSeek-V4-Pro-0813 now on @nebiustf pushing intelligence while being pretty economical! viz from Token Factory model visualizer : sujee.github.io/practical-llm-…

Nebius Token Factory @nebiustf

2 weeks ago

Two new models for coding and agent workflows are now live on Nebius Token Factory. DeepSeek V4-Pro-0813 is the official V4 Pro release, built for coding agents that use tools, reason through complex problems, and work across multiple steps. GLM-5.3 brings Z. ai’s latest

14 6 78 16K 7

2 0 6 367 1

View Details

Sujee Maniyam @sujee_dev

2 weeks ago

As I was updating my model indexes, I noticed that @ArtificialAnlys has updated its Intelligence Index. The new AA Intelligence Index v4.3 introduces several significant changes, including a tougher set of evaluations. You can read about the changes here: x.com/ArtificialAnly… One immediate effect: scores are lower across the board. That doesn't necessarily mean the models got worse. The benchmark suite got harder, so the scale has effectively shifted. At the top of the latest leaderboard: - Claude Fable 5.1: 53 = 66 → 53 = ▼ 13 - GPT-6 Astra: 53 = 61 → 53 = ▼ 8 And the top 3 open-weight models: 1. GLM-5.3: 45 2. Kimi K3: 44 (60 → 44 = ▼ 16) 3. GLM-5.3-Flash: 42 (57 → 42 = ▼ 15 ) The score changes for open models are pretty substantial. See the images below 👇 for the changes and the latest leaderboard. As they say, it's all relative 😀 What matters most is how models compare against each other under the same evaluation suite.

Artificial Analysis @ArtificialAnlys

3 weeks ago

Announcing Artificial Analysis Intelligence Index v4.3, upgrading Terminal-Bench to 4.0 and adding AutomationBench-AA, an agentic workflow automation benchmark with a private test set. This is a continuation of our rollout of Intelligence Index v5 Changelog (Index v4.2 → Index

190 159 2K 1.1M 369

0 0 1 191 0

View Details

Sujee Maniyam @sujee_dev

2 weeks ago

@ArtificialAnlys I was just refreshing my open-model rankings and saw the shift in scores Kimi K3: 60 → 44 GLM-5.3-Flash: 57 → 42 GLM-5.2: 53 → 39 A good reminder that the absolute number matters less than the relative ranking under the same benchmark suite.

0 0 0 36 0

View Details

Sujee Maniyam @sujee_dev

2 weeks ago

Had a fun time geeking out and building with @dN0t and @devopsjacquie - using open models for coding -- they are surprisingly good (and cheaper) - using Hermes to find vintage auto parts! - new models in @nebiustf Get started by singing up to the Builder Program : lnkd.in/gkUXHpMS And share with us what you build. x.com/i/broadcasts/1…

0 1 6 152 0

View Details

Nebius @nebiusai

2 weeks ago

The Nebius AI Builder Program is now available. AI isn’t just a model you call anymore. It’s a system you build. And builders, not a few closed labs, will decide what it becomes. The open ecosystem has all the pieces. We want to make it easier to put them together and start building. The program is free, with $400+ in credits and discounts, working code and cookbooks, office hours with engineers, and a community to build with. We’re joined by @NVIDIAAI, @LangChain, @huggingface, @cognition, @OpenHandsDev, @tavilyai, @TolokaAI, @composio, @PrimeIntellect, @MiniMax_AI, @Alibaba_Qwen, and more joining soon. Join with the link in the comments 👇

476 392 1K 2.6M 1K

View Details

Sujee Maniyam @sujee_dev

2 weeks ago

sujee.dev/post/fine-tuni…

0 0 0 5 0

View Details

Sujee Maniyam @sujee_dev

3 weeks ago

Here’s a quick explainer on fine-tuning, distillation, and quantization - using LEGO® bricks 🧱 🟢 Fine-tuning: specialize a pretrained model with new data. 🟠 Distillation: train a student model to learn from a teacher. 🟡 Quantization: use lower-precision weights to reduce memory and bandwidth use. Video below 👇 More detail: sujee.dev/post/fine-tuni… What should I explain next with LEGO: LoRA, pruning, or speculative decoding? Drop your vote in the comments 👇

1 1 6 196 0

View Details

Sujee Maniyam @sujee_dev

3 weeks ago

Went to buy a couple of 4TB portable SSDs and almost fell off my chair looking at the prices 😳 Damn. See screenshots 👇 @demian_ai has been writing about the memory/storage crunch driven by the AI boom. It is finally sinking in for me :-) A few of his posts: - x.com/demian_ai/stat… - x.com/demian_ai/stat… - x.com/demian_ai/stat…

dylan ツ @demian_ai

2 months ago

$PENG turned scarcity into revenue. Samsung made memory stretch. Micron locked down wafers. Korea broke the wrapper. this week, the book looked flat but the stack was being repriced ↓ Under a flat headline, the market made a clear choice: pay for what can turn scarcity into