|
Kimi-K3 on HuggingFace
|
1,378 |
-- |
2026-07-27 |
|
Show HN: Sweep, Open-weights 1.5B model for next-edit autocomplete
|
530 |
-- |
2026-01-21 |
|
Anatomy of a Frontier Lab Agent Intrusion: A Timeline of the July …
|
466 |
-- |
2026-07-28 |
|
Kimi K2.7-Code: open-source coding model with better token efficiency
|
455 |
-- |
2026-06-12 |
|
Show HN: Hacker News archive (47M+ items, 11.6GB) as Parquet, updated every …
|
399 |
-- |
2026-03-14 |
|
GLM-4.7-Flash
|
371 |
-- |
2026-01-19 |
|
Inflect-Micro-v2: complete voice in 9.36M parameters
|
213 |
-- |
2026-07-26 |
|
DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
|
159 |
-- |
2026-04-24 |
|
Show HN: Text-to-video model from scratch (2 brothers, 2 years, 2B params)
|
156 |
-- |
2026-01-22 |
|
Waypoint-1: Real-Time Interactive Video Diffusion from Overworld
|
92 |
-- |
2026-01-23 |
|
Show HN: 30k IKEA items in flat text
|
55 |
-- |
2026-01-07 |
|
Anyone Can Clone Your Voice Now
|
45 |
-- |
2026-01-26 |
|
Qwen/Qwen3.6-27B · Hugging Face
|
41 |
-- |
2026-04-22 |
|
Continuous batching (2025)
|
39 |
-- |
2026-02-15 |
|
Anatomy of BoltzGen
|
31 |
-- |
2026-01-04 |
|
DeepSeek-V4 Technical Report [pdf]
|
26 |
-- |
2026-04-24 |
|
Gemma 4 E2B running in-browser at 255 tok/s
|
19 |
-- |
2026-06-17 |
|
Qwen 3.5 small models out
|
18 |
-- |
2026-02-24 |
|
Netflix just dropped their first public model on Hugging Face: VOID
|
18 |
-- |
2026-04-04 |
|
Hugging Face Storage Buckets: Mutable, non-versioned object storage at $12/TB
|
18 |
-- |
2026-03-10 |
|
Rio 3.5 Open 397B – from Rio de Janeiro's city government
|
17 |
-- |
2026-06-13 |
|
Google translategemma 4B Translation Models
|
16 |
-- |
2026-01-19 |
|
Z.ai GLM 5.2
|
16 |
-- |
2026-06-16 |
|
Flux.2 Klein 4B (Apache 2.0)
|
14 |
-- |
2026-01-19 |
|
DeepSeek V4 Flash
|
14 |
-- |
2026-04-24 |
|
Show HN: 17MB model beats human experts at pronunciation scoring
|
13 |
-- |
2026-02-20 |
|
DeepSeek-V4: a million-token context that agents can use
|
13 |
-- |
2026-04-28 |
|
Distilling 100B+ Models 40x Faster with TRL
|
13 |
-- |
2026-04-12 |
|
Qwen 3.5
|
12 |
-- |
2026-02-16 |
|
Minimax M2.7 Weights Released
|
11 |
-- |
2026-04-12 |
|
5.6x throughput on Kimi K2.6 by speculating less
|
11 |
-- |
2026-04-21 |
|
Kimi K2.6
|
11 |
-- |
2026-04-20 |
|
Tencent/Hy3: 295B MoE model rivals trillion scale SOTA
|
11 |
-- |
2026-07-09 |
|
VibeVoice-ASR: speech-to-text model designed to handle 60-minute long-form audio
|
11 |
-- |
2026-01-31 |
|
Qwen3.5 Small: 0.8B, 2B, 4B, 9B Released
|
10 |
-- |
2026-03-02 |
|
Show HN: Trained an LLM to predict "What will Trump do?"
|
10 |
-- |
2026-02-20 |
|
Nvidia Nemotron 3-Nano 30B-A3B-BF16
|
10 |
-- |
2026-01-31 |
|
GLM-5.2: Built for Long-Horizon Tasks
|
10 |
-- |
2026-06-17 |
|
MiniMax-M3: A native multimodal model with 1M context
|
10 |
-- |
2026-06-12 |
|
Alibaba open-sources Qwen3.6-35B-A3B, a 35B MoE model with 3B active parameters
|
10 |
-- |
2026-04-16 |
|
Xiaomi MiMo-v2.5-Pro Open-Sourced: 1T Parameter Model
|
9 |
-- |
2026-04-28 |
|
Soofi: European sovereign LLM trained in 2 months
|
9 |
-- |
2026-07-11 |
|
Nvidia releases 8B model with learned 8x KV cache compression
|
9 |
-- |
2026-01-24 |
|
DeepSeek: Thinking with Visual Primitives [pdf]
|
9 |
-- |
2026-04-30 |
|
Mistralai/Leanstral-1.5-119B-A6B
|
8 |
-- |
2026-07-03 |
|
Reka Edge – 7B fast, efficient VLM (open-weights)
|
8 |
-- |
2026-03-11 |
|
Lance – native supports image and video understanding, generation, and editing
|
8 |
-- |
2026-05-24 |
|
Bonsai 1.7B in the browser: a 290MB 1-bit LLM on WebGPU
|
8 |
-- |
2026-04-16 |
|
Gemma 4 running client-side in WebGPU
|
8 |
-- |
2026-04-04 |
|
Show HN: Hacker News RSS Feed Directory
|
8 |
-- |
2026-04-04 |
|
Show HN: Marlin-2B: a tiny VLM to extract structured information from videos
|
7 |
-- |
2026-05-18 |
|
Zyphra releases the ZAYA1-8B MoE model optimized for intelligence density
|
7 |
-- |
2026-05-06 |
|
Carbon: Autoregressive Genomic Foundation Model
|
7 |
-- |
2026-05-19 |
|
Show HN: We beat Gemini Embedding 2 by training only 16M params …
|
7 |
-- |
2026-07-10 |
|
Nex N2 Pro: Frontier agentic performance at 400B
|
7 |
-- |
2026-06-08 |
|
The ultimate guide to RL environments: building and scaling them in the …
|
7 |
-- |
2026-05-05 |
|
DeepSeek-V4
|
7 |
-- |
2026-04-24 |
|
We OCR'ed 30k papers using Codex, open OCR models and Jobs
|
7 |
-- |
2026-04-21 |
|
Nemotron 3 Nano 4B: A Compact Hybrid Model for Efficient Local AI
|
7 |
-- |
2026-03-18 |
|
Show HN: Open-source LLM and dataset for sports forecasting (Pro Golf)
|
7 |
-- |
2026-02-24 |
|
Datadog scales time series foundation models to 2.5B parameters
|
6 |
-- |
2026-05-14 |
|
Gemma 4 Uncensored (autoresearch results)
|
6 |
-- |
2026-04-05 |
|
Show HN: A 0.3B model that redacts PII in all 24 EU …
|
6 |
-- |
2026-05-13 |
|
Minimax M3 weights to be released on Friday
|
6 |
-- |
2026-06-11 |
|
Arcee-AI Trinity-Large-Thinking
|
6 |
-- |
2026-04-02 |
|
Show HN: Nanbeige 4.1-3B running in the browser via WebGPU
|
6 |
-- |
2026-02-19 |
|
Show HN: A 150M model that extracts verbatim evidence spans for RAG, …
|
6 |
-- |
2026-06-10 |
|
"Optimal Cognitive Core"- specialized 1.7B model for grounded question answering
|
6 |
-- |
2026-06-03 |
|
What the Community Is Running
|
6 |
-- |
2026-05-31 |
|
Show HN: We trained a 32B model to beat Opus 4 at …
|
6 |
-- |
2026-04-20 |
|
Show HN: Vocab extractor for language learners using Stanza and frequency ranks
|
6 |
-- |
2026-03-29 |
|
The Future of the Global Open-Source AI Ecosystem: From DeepSeek to AI+
|
6 |
-- |
2026-02-07 |
|
First DeepSeek V4 Flash-Base-Int4 Quant
|
5 |
-- |
2026-04-27 |
|
APIEval-20: A Benchmark for Black-Box API Test Suite Generation
|
5 |
-- |
2026-03-30 |
|
Show HN: Voice gender classifier for European voice AI (1MB, ONNX, 4ms)
|
5 |
-- |
2026-05-12 |
|
Don't Train the Model, Evolve the Harness
|
5 |
-- |
2026-07-03 |
|
PP-OCRv6
|
5 |
-- |
2026-06-16 |
|
I made a CPU only spiking neuron network lib that comes pretty …
|
5 |
-- |
2026-05-30 |
|
Gemma 4 Models
|
5 |
-- |
2026-04-02 |
|
Do Bubbles Form When AIs Simulate Capitalism?
|
5 |
-- |
2026-02-24 |
|
Craft – image models can think like LLMs
|
5 |
-- |
2026-02-06 |
|
Sopro v1.5: A 135M TTS model trained for ~$100, runs 20× real-time …
|
5 |
-- |
2026-02-05 |
|
Kernels: Major Updates
|
5 |
-- |
2026-07-06 |
|
ThinkingCap-Qwen3.6-27B: Qwen3.6 capabilities with 50% fewer thinking tokens
|
5 |
-- |
2026-07-06 |
|
Step 3.7 Flash – 198B-A11B MoE vision-language model
|
5 |
-- |
2026-05-30 |
|
ZooL4nD3r: Translate a passage across 961 learned discourse communities
|
5 |
-- |
2026-05-06 |
|
Xiaomi open-sources MiMo-V2.5: 311B A15B 1M-context omnimodal model
|
5 |
-- |
2026-04-28 |
|
Local Agents with Llama.cpp and Pi
|
5 |
-- |
2026-03-12 |
|
PRX Part 3 – Training a Text-to-Image Model in 24h
|
5 |
-- |
2026-03-03 |
|
Qwen3.5
|
5 |
-- |
2026-02-16 |
|
GLM-Image
|
5 |
-- |
2026-01-14 |
|
Show HN: Forecasting my backyard weather with a 22M time-series model
|
4 |
-- |
2026-05-17 |
|
Amalia – an open-source language model targeting European Portuguese
|
4 |
-- |
2026-07-01 |
|
Empero-AI/Qwythos-9B-Claude-Mythos-5-1M
|
4 |
-- |
2026-06-29 |
|
Show HN: Mate – Emotional layer on top of LLMs
|
4 |
-- |
2026-04-03 |
|
Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled
|
4 |
-- |
2026-03-03 |
|
Show HN: EPEE, an expert-annotated ASL dataset from native Deaf signers
|
4 |
-- |
2026-07-09 |
|
ZeroLabs – 100x cheaper than ElevenLabs (free forever locally) with open models
|
4 |
-- |
2026-07-02 |
|
Reachy Mini bot goes local
|
4 |
-- |
2026-05-28 |
|
Mistral Medium 3.5 128B
|
4 |
-- |
2026-04-30 |
|
A long-term memory system that turns agent trajectories into reusable guidelines
|
4 |
-- |
2026-04-10 |
|
GPT-OSS-20B-Vision: First Community VLM for GPT-OSS, Trained on a DGX Spark
|
4 |
-- |
2026-02-19 |
|
DeepSeek OCR2
|
4 |
-- |
2026-01-27 |
|
Show HN: I trained a 1B LLM from scratch for $315 and …
|
4 |
-- |
2026-07-02 |
|
Ai2 ACE2S – Simulate atmospheric variability – Scale of days to centuries
|
4 |
-- |
2026-06-16 |
|
Why do LLM outputs get worse even when metrics stay stable? [pdf]
|
4 |
-- |
2026-05-07 |
|
AI evals are becoming the new compute bottleneck
|
4 |
-- |
2026-05-04 |
|
OpenAI Privacy Filter
|
4 |
-- |
2026-04-22 |
|
"Darwin-27B-Opus: Surpassing the Foundation Model Without Training"
|
4 |
-- |
2026-04-13 |
|
Liberate Your OpenClaw
|
4 |
-- |
2026-04-04 |
|
Jane Street – neural net puzzle
|
4 |
-- |
2026-04-01 |
|
Show HN: ArXiv metadata as Parquet files (2.99M papers, 1.44GB, 417 files)
|
4 |
-- |
2026-03-24 |
|
Deploying Open Source Vision Language Models (VLM) on Jetson
|
4 |
-- |
2026-02-25 |
|
Custom Kernels for All from Codex and Claude
|
4 |
-- |
2026-02-18 |
|
PersonaPlex-7B: full-duplex voice model that listens and talks at the same time
|
4 |
-- |
2026-02-15 |
|
MiniMax M2.5 weights
|
4 |
-- |
2026-02-13 |
|
One Year Since the "DeepSeek Moment"
|
4 |
-- |
2026-01-27 |
|
DeepSeek MHC: Manifold-Constrained Hyper-Connections
|
4 |
-- |
2026-01-02 |
|
Show HN: Open dataset for distilling GPU audio2face models to CPU students
|
3 |
-- |
2026-06-09 |
|
Two Years of Local AI on a Laptop: When Open Models Outpaced …
|
3 |
-- |
2026-05-11 |
|
HuggingFace launches pre-compiled machine learning kernels repository
|
3 |
-- |
2026-04-15 |
|
Show HN: Prompt injection detector beats ProtectAI by 19% accuracy, 8.9x smaller
|
3 |
-- |
2026-04-08 |
|
Reproducing Anthropic's "Counting Manifold"
|
3 |
-- |
2026-02-20 |
|
TabFM: Zero-shot tabular foundation model from Google Research
|
3 |
-- |
2026-07-04 |
|
ML research datasets from ArXiv and Semantic Scholar (JSONL, quality-scored)
|
3 |
-- |
2026-06-16 |
|
A Blockade on Knowledge. After June 12, Nothing Will Be the Same
|
3 |
-- |
2026-06-14 |
|
Deepfake Detector Robustness Testing
|
3 |
-- |
2026-06-07 |
|
QVAC MedPsy: Medical and Healthcare Models for Edge Device
|
3 |
-- |
2026-05-07 |
|
Hugging Face moves safetensors to the PyTorch Foundation
|
3 |
-- |
2026-04-08 |
|
Training mRNA Language Models Across 25 Species for $165
|
3 |
-- |
2026-04-07 |
|
Hugging Face Provides an S3 Alternative
|
3 |
-- |
2026-03-24 |
|
SynthVision: Building a 110K Synthetic Medical VQA Dataset
|
3 |
-- |
2026-03-23 |
|
QORA-LLM-2B – Pure Rust ternary inference, no multiplication needed
|
3 |
-- |
2026-03-11 |
|
Pure Rust, zero dependencies AI models, runs locally, free forever
|
3 |
-- |
2026-02-28 |
|
TranslateGemma now runs 100% in the browser on WebGPU with Transformers.js v4
|
3 |
-- |
2026-02-25 |
|
Multi-vector Grep for code agents, save 15% tokens, local
|
3 |
-- |
2026-02-12 |
|
Step 3.5 Flash LLM model, agentic coding ~18x faster than GLM 4.7 …
|
3 |
-- |
2026-02-02 |
|
Introduction to Reinforcement Learning and Its Role in LLMs
|
3 |
-- |
2026-07-11 |
|
Native-speed vLLM transformers modeling back end
|
3 |
-- |
2026-07-08 |
|
IOL-AI 2026 Challenge: Can Your Model Solve Linguistics Olympiad Problems?
|
3 |
-- |
2026-07-06 |
|
Filter Hugging Face Models Page by Hardware
|
3 |
-- |
2026-06-30 |
|
MFM: PINN based Motion Foundation Model
|
3 |
-- |
2026-06-29 |
|
HRM-Text: Efficient Pretraining Beyond Scaling
|
3 |
-- |
2026-06-25 |
|
The first open, vision-driven real-time interaction model
|
3 |
-- |
2026-06-23 |
|
Show HN: Multi-Agent AI trading firm simulation powered by small language models
|
3 |
-- |
2026-06-20 |
|
Genoma Labs' open 14B agentic coding model trained on Kraken
|
3 |
-- |
2026-06-17 |
|
Donate Agent Traces
|
3 |
-- |
2026-06-15 |
|
Unknown Is Not False: An AI Agent Pre-Execution Checklist
|
3 |
-- |
2026-06-11 |
|
Nvidia/Nvidia-Nemotron-3-Ultra-550B-A55B-BF16
|
3 |
-- |
2026-06-08 |
|
New SoTA open source TTS model from Boson AI
|
3 |
-- |
2026-06-05 |
|
Aura: Action-Gated Memory for Robot Policies at Constant VRAM
|
3 |
-- |
2026-06-04 |
|
Direct Preference Optimization Beyond Chatbots
|
3 |
-- |
2026-06-04 |
|
Gemma 4 12B appears in Hugging Face
|
3 |
-- |
2026-06-03 |
|
Nvidia: Nemotron Labs Diffusion 14B
|
3 |
-- |
2026-05-23 |
|
Karpathy's autoresearch, 50 DPO experiments, 300 human judges
|
3 |
-- |
2026-05-21 |
|
The Open Agent Leaderboard
|
3 |
-- |
2026-05-18 |
|
Physics-intern: an autonomous agentic framework for physics research
|
3 |
-- |
2026-05-12 |
|
Cuarzo-100K v2 – Python↔EN/ES/FR/ZH, 100% AST verified across all 4 languages
|
3 |
-- |
2026-05-11 |
|
Show HN: Open Source FreeCAD dataset for CAD generation tasks
|
3 |
-- |
2026-05-08 |
|
Building a Fast Multilingual OCR Model with Synthetic Data
|
3 |
-- |
2026-04-28 |
|
Neo-Unify: An Encoder-Free, Native Multimodal Paradigm (SenseTime)
|
3 |
-- |
2026-04-14 |
|
The Open-Source Recipe for Teaching a Robot to Fold Your Clothes
|
3 |
-- |
2026-04-08 |
|
Nemotron OCR v2 by NVIDIA
|
3 |
-- |
2026-04-03 |
|
Granite 4.0 3B Vision: Compact Multimodal Intelligence for Enterprise Documents
|
3 |
-- |
2026-04-01 |
|
TRL v1.0: Post-Training Library Built to Move with the Field
|
3 |
-- |
2026-04-01 |
|
Mr. Chatterbox is an LLM trained exclusively on Victorian-era British texts
|
3 |
-- |
2026-03-29 |
|
CanViT: Toward Active-Vision Foundation Models
|
3 |
-- |
2026-03-25 |
|
Build a Domain-Specific Embedding Model in Under a Day
|
3 |
-- |
2026-03-23 |
|
Show HN: 518K Vietnamese legal documents (1924–2026)
|
3 |
-- |
2026-03-22 |
|
Keep the Tokens Flowing: Lessons from 16 Open-Source RL Libraries
|
3 |
-- |
2026-03-21 |
|
9B parameter coding agent model fine-tuned on top of Qwen3.5-9B
|
3 |
-- |
2026-03-13 |
|
Leaderboard of Leaderboards – A Real-Time Meta-Ranking of AI Benchmarks
|
3 |
-- |
2026-03-13 |
|
Smol AI WorldCup: What Small LLMs Can Do
|
3 |
-- |
2026-03-10 |
|
The Synthetic Data Playbook: Generating Trillions of the Finest Tokens
|
3 |
-- |
2026-03-08 |
|
Modular Diffusers – Composable Building Blocks for Diffusion Pipelines
|
3 |
-- |
2026-03-06 |
|
Phi-4-reasoning-vision-15B
|
3 |
-- |
2026-03-05 |
|
DataClaw
|
3 |
-- |
2026-02-25 |
|
GGML and llama.cpp join HF to ensure the long-term progress of Local …
|
3 |
-- |
2026-02-20 |
|
GLM-5: From Vibe Coding to Agentic Engineering
|
3 |
-- |
2026-02-18 |
|
Show HN: DeepBrainz-R1 – Reasoning-First Small Models for Agentic Systems
|
3 |
-- |
2026-02-05 |
|
Qwen3-coder-next: SOTA open source coding model
|
3 |
-- |
2026-02-03 |
|
We got Claude to teach open models how to write CUDA kernels
|
3 |
-- |
2026-01-29 |
|
Architectural Choices in China's Open-Source AI Ecosystem
|
3 |
-- |
2026-01-27 |
|
Bridging the Gap Between AI Agent Benchmarks and Industrial Reality
|
3 |
-- |
2026-01-21 |
|
Jupyter Agents: training LLMs to reason with notebooks
|
3 |
-- |
2026-01-11 |
|
Nvidia brings agents to life with DGX Spark and Reachy Mini
|
3 |
-- |
2026-01-06 |
|
Falcon-H1-Arabic: Pushing the Boundaries of Arabic Language AI
|
3 |
-- |
2026-01-06 |
