klein-ml have finished stage 1 of KV1 1.2B, we don't think we will need additional stages.
KV1 2.6B is near the midpoint of training, going well so far, its going to be our most agentic model yet.
We have released Klein Code Bench klein-ml/klein-code-bench, a benchmark testing whether models can build complete projects from scratch. It contains a easy, intermediate, hard and max subset.
Here are some benchmarks of the current KV1 1.2B
It retains 75.8% of FP16 intelligence.
We are also announcing that we at
klein-ml are working on Klein 24B A2B which will be our first model featuring ternary weight quantization.
We're going to take some models, do a bit of continued training on them and QAT them for only a 2 bit KV cache. And because of what happened with BGA, and this is also kind of some research stuff, here's what we're going to change: We're going to always, before we release how and make any claims, release our results. We also would like (idk if we can) to have some external reviewers before releasing them, I'M talking to YOU @AtAndDev and @mithulaartigala And we're going to properly cite other research.
The models we are going to release is: KV1 1.2B, based on LFM 2.5 1.2B Instruct KV1 2.6B, based on LFM 2.5 2.6B Instruct KV1 4B, based on Qwen3.5 4B.
If we get enough compute, we may also train: KV1 9B
And if you guys want to a ~27B model, you guys decide which, either on LFM 2.5 24B A2B, Qwen3.8 27B, or Qwen3.6 35B A3B.
Our goal is to get 2 bit, at 4 bit quality.
Here's an image of our current progress, KV1 is at 3 bit, not 4 bit, but it will improve!
And we're releasing something major again now, if you have seen Grok Bot, its very powerful, but there isn't really a local alternative yet.
Today we're releasing Minds, they're personal AI bots that live on your computer. They can run anywhere, on your phone, on your mac, on windows, and even on linux.
They support OpenAI Oauth as the main integration because most people already have a ChatGPT sub.
But locally, It supports ANY openai compatible api which allows you to setup ollama and llama.cpp and everything else.
It uses Jev for command approval on OpenRouter and Muse Spark Transcribe for transcription.
1. Should I release all Pebble 1.5 models together in one big launch, or release each model individually as it's ready? 2. What would you like to see from the Pebble models? Any specific features, capabilities, or improvements? 3. What would you like to see from basically AI as a company in general?
We are now announcing BananaMind 3 🍌! (not ai for those emoji guys)
All models will use BGA (which is almost just NSA) and our BM3X architecture. Its sizes will be: BananaMind 3 Flash Lite, 3M parameters at a context of 8K context. BananaMind 3 Lite, 10M parameters with 16K context. BananaMind 3 Flash, 25M Parameters with 16K context. BananaMind 3 Pro, 50M parameters with 24K context. BananaMind 3 Ultra, 100M parameters with 32K context. And lastly, BananaMind 3 Max with 150M parameters and 64K CONTEXT.
I can assure you BananaMind 3 Max WILL beat GPT X3 or match it, we won't release it otherwise. We hope for a 40+ INTELLIGENCE INDEX!
BananaMind 3 may also be partnered with dot labs.
We will cancel BananaMind 2.1 and BananaMind 2 Ultra.
As of the BETU SLM Leaderboard we may need to release it after October 11, im very busy right now (even though we said We will release BETU leaderboard before Oct 11 😟)
We have released BGA! And wow, It provides 256x (and 512x at the end of 1M) yes 256x LESS attention compute at 1M context window. That means you can train a 1M context window at the compute of a ~4K context window.
Hi! We're right now experimenting with so many new architectures!
We are also going to release BananaMind Gate Attention very soon, its a new attention that can make attention 128x cheaper at 1M Context! Check my account for HF blogs it will release there!
We're introducing ACR 1.0. We trained this model on a 5070 Ti for weeks, here are some of the architecture details: 57M parameters, with one M and one G stream. When we tested it on benchmarks, we got these results: Benchmark Full G-Only Delta PIQA 62.24% 53.43% +8.81 ARC-Easy 41.96% 32.28% +9.68 HellaSwag 33.19% 29.08% +4.11 Tiny ToM 40.65% 33.75% +6.90 ArithMark 3.0 33.40% 32.80% +0.60 Base Bench 1.1 51.71% 40.29% +11.42
We will release the BEST SLM Leaderboard before Oct 11.
It will feature everything: Easy to use model picker. EXTREMELY Easy way to add your own models (2 click) MULTIPLE leaderboard for different model types
Hey everyone! I got sidetracked from my main projects and decided to test out the BananaAll app and see if I could make a small model not regress too much during SFT. Here is what happened:
- Thank you to @Banaxi-Tech for the BananaAll app (works perfectly on Windows and CPU) - GPT-6 Sol for knowing how to merge some confusing files created by the app - Myself for the idea - Someone else somewhere who might have contributed to some of my ideas and might in the future - And readers like you!
We're releasing a MAJOR update to the BananaAll SLM Super App. If you want to use a custom architecture, previously you had to go trough reviewing the code yourself, now add an Openrouter API key and review it with GPT 6 Luna in one button. A review cost be half a cent so anyone can try it. This is one of the main features. Now ROCm, AMD and Windows, Mac support. Colab and Molab support.
Detailed list of features: Get improved Windows Python detection and support paths for compatible AMD ROCm, Intel XPU, and Apple MPS setups. Choose local training or export a self-contained Python script for Colab or Molab. Notebook runs produce a downloadable model ZIP. Start pretraining with an existing model’s tokenizer, or train a new one from your datasets. Try experimental 1.58-bit Ternary fake-quantized training on NVIDIA GPUs. Watch live tokens per second. Model compilation is on by default and falls back automatically if it fails. Build custom architectures with separate configuration and modeling files, then review the training code manually or with optional OpenRouter AI Review. Install from source with the new coding-agent instructions. This release also fixes inflated loss reporting for custom models.
And for those users who didn't want to try it out just because installation would be so hard, it isnt now. Go to any coding agent (Pi, Claude Code, Codex, OpenCode, basically all work), and just paste "Install BananaAll for me. Fetch and follow https://raw.githubusercontent.com/BananaMind/BananaAll/main/agent_install.txt." That's it.
auto 200m 2 is out now! Auto is a series of models designed to classify if a tool called by an AI agent is safe to run or not, like how the "auto" approval modes work in codex or claude. This is the smallest model in the series by far and it still packs a punch! Over 96 percent accuracy on the held out benchmark, beating deepseek v4 0731 and absolutely destroying regex at classifying if a tool call is safe to run (to be fair it is a bit of a weird and very hard benchmark but it works!) check it out: ProCreations/auto-200m-2
We're excited to release BananaAll, our SLM Super App.
It allows you to do EVERYTHING you need to do to trains SLMs in a single app, no terminal, no 30 chrome tabs.
The train tab allows you to train models, select datasets from presets, and use other ones with auto mapping, model size slider, it automatically generates a training script for you.
Then after you've trained the model or want to compare it to competitors, the evaluation tab, run ARC EASY, ARC Challenge, Hellaswag, PIQA, Arithmark 3, BananaMind Base Bench and more! Simple Results screen.
And lastly the inference tab, run your trained models or others.
Normally you would need seperate apps or scripts for that, but the BananaAll Super App lets you do all of that in a single app.
We also trained a small 2.5M parameter model on 200M tokens of Fineweb edu, The results: BananaMind Base Bench 854 and 53% on PIQA. On only 200M tokens.
We're working on Pebble 1.5. Here's what we know so far:
- They will be better than the last generation. 99.99% certain. - Expanded context lengths of at least 16,384 tokens, with the flagship potentially reaching 32,768. - A Mamba3-based architecture with some other new architectural designs we're experimenting with. - Native CPU compatibility — something we failed at with the last generation. - Natively multilingual and multimodal???
2. SmolCodeBench
A code benchmark designed specifically for small models, because there really isn't a good one right now.
3. SENTRY
VOID is working on something called SENTRY — System for Evaluating Neural Threats, Responses, and Yields.
More on that soon.
4. basically OS
It's an operating system/app/harness. We're still deciding.
5. Finances
Trying to balance the finances after purchasing a Hugging Face Pro subscription.
hi everyone we have released nacr its not just any model, its nacr we have 6 more features and this model only uses 20% of its total capacity! check it out at saicr/nacr we're currently working on expanding access as we do more research but right now you have to use our gated access form
follow
saicr if you're interested if you want to join saicr, first read the entire nacr readme, then press the join button.
klein-ml are announcing KleinMem
