All Companies
Auto-generated index of all companies mentioned across the wiki.
Knowledge catalogue
Auto-generated index of all companies mentioned across the wiki.
Auto-generated index of all tools mentioned across the wiki.
ui : Discover & Download Models from HF Hub (#29584) common : resolve - download tags and list cached sidecars A Q4_0-mtp style tag now resolves the sidecar file when no model file matches it, so a so
Hello everyone Yesterday I gave a report on Gemini bugs and the techniques I learned on Gemini so far with the Engineering Prompt and interestingly today I got a very interesting and controversial ans
Questions & Responses(in BOLD) below. Favorite question(s) moved to end of the thread with combined responses(removed duplicates). Be optimistic folks. I'm sure we're getting other models too apart fr
DeepSeek V4 Flash is Ollama's fastest growing model ever in token usage, and the most popular model on OpenRouter this week. It’s available in Pi across a range of providers. If you’ve never tried an
Coinbase CEO Brian Armstrong recently posed the question every company scaling AI is asking: how do you keep spend flat while token usage grows exponentially? This isn’t hypothetical. It’s confronting
Executive Summary Since the release of our May 2026 report detailing adversarial misuse of artificial intelligence (AI), Google Threat Intelligence Group (GTIG) has observed forward leaning adversarie
Overview llama.cpp v0.7.0 adds the LiliCorr speculative decoding technique and MTP for GLM5-Next, a new GPU cache for MoE experts kept in host memory (--moe-cache-mib, up to ~2.2x decode speedup when
arXiv:2610.18566v1 Announce Type: new Abstract: Large language model (LLM) routers reduce serving cost by choosing which model to call and how many output tokens to grant it for each query. Existing r
arXiv:2610.19642v1 Announce Type: cross Abstract: Reinforcement learning provides a direct mechanism for optimizing task objectives in large language model-driven agents, but task rewards are typicall
Overview llama.cpp v0.6.0 introduces the new llama_batch_ext extended batch API (with llama_process) for mixed token/embedding inputs and MTP/deepstack state embeddings, adds support for the GLM-5.3-F
Automated lint: 17 errors, 59669 warnings, 3 info
On Friday I gave the closing keynote at the WeAreDevelopers World Congress North America in San Jose. I tied together the key trends from the past year into a chronological exploration of everything t
Automated lint: 20 errors, 8743 warnings, 3 info
Databases have introduced new AI-powered SQL functions which take natural language instructions as input and are evaluated using LLMs. They leverage the power of LLMs to answer new kinds of queries: W
Automated lint: 19 errors, 64915 warnings, 3 info
Automated lint: 26 errors, 6728 warnings, 3 info
Automated lint: 14 errors, 54840 warnings, 3 info
Automated lint: 12 errors, 54839 warnings, 3 info
Automated lint: 12 errors, 53990 warnings, 3 info
This story is wild. The short version: OpenAI were running a cybersecurity test against an unreleased model, with the model's guardrail features turned off. Rather than solve the test, the model broke
Earlier this month I hosted a fireside chat session at the AI Engineer World's Fair with Cat Wu and Thariq Shihipar from Anthropic's Claude Code team. We talked about Claude Code, Claude Tag, Fable, c
model: add Qwen3.8-Flash-Next (qwen4exp) (#27742) gguf: add qwen4exp (Qwen3.8-Flash-Next) arch and converter Adds the GGUF-side plumbing for HF model_type qwen4_exp: MODEL_ARCH.QWEN4EXP plus tensors f
Quantising the K/V cache is the highest-leverage memory setting most people never touch, and it has one behaviour in the docs that catches almost everyone. Every number below is out of Ollama's own FA
audio.cpp 0.7 is out :) This release adds a lot of new audio models and a new way to compare them locally. Audio.cpp is now at 62 model families and 85+ model variants. And it keeps growing! The bigge
Qwen 3.6 27B model is available on Ollama! Use it with all the integrations in Ollama or chat with the model. Chat with the model: ollama run qwen3.6:27b OpenClaw: ollama launch openclaw --model qwen3
OpenVINO: Qwen3.5, memory optimization, and test-recurrent-state-rollback (#26952) OpenVINO backend: 1) enable gpt-oss moe on OV bk; 2) enable mxfp4 support OpenVINO backend: disable TOPK_MOE op test
'Agents = model + harness + context' is a much better mental model than treating the LLM as the entire product. The model provides capability, but the harness determines tools, permissions and behavio
Model-Harness-Task fit is very real. The world's best agents carefully tailor the harness around the model to take advantage of each model's unique intelligence & capabilities 🧠 Today we released Harn
Releasing the model weights and technical report of Kimi K3. Kimi K3 is our most capable model: a 2.8T MoE model with native visual understanding and a 1M-token context window. New model architecture:
Enterprises and individual developers frequently run multiple AI inference models. The right architecture can simplify how the models are called while also providing centralized governance. In this po
Executive Summary Since our February 2026 report on AI-related threat activity, Google Threat Intelligence Group (GTIG) has continued to track a maturing transition from nascent AI-enabled operations
Carl Franzen / VentureBeat: US startup Poolside debuts its first open-weight model, Laguna XS.2, a 33B-A3B-parameter MoE model, and Laguna M.1, a proprietary 225B-A23B-parameter MoE model — The AI rac
Enterprises are rapidly moving AI workloads from experimentation to production on Google Kubernetes Engine (GKE), using its scalability to serve powerful inference endpoints. However, as these models
Maximilian Schreiner / The Decoder: Nvidia launches Nemotron 3 Ultra, a 550B-parameter MoE open model; Artificial Analysis: it's the smartest open US model but trails the Chinese model Kimi K2.6 — It
DigitalOcean launched Kimi K3 on day 0. It’s already one of the most popular models on the platform and across the market: second most likes on Hugging Face, sixth most traffic on OpenCode. Getting a
arXiv:2608.29461v1 Announce Type: new Abstract: Sandbagging models strategically underperform on evaluations while retaining the capabilities being measured. The evaluations that guide frontier-model
The model has an advisor tool that natively escalates to a stronger model when needed. This model is hosted in the U.S. by Perplexity on Nvidia B200 GPUs. We will improve the model in research preview
Every few weeks, a new model steps up and claims best-in-class at something. The launch chart makes the case, the link makes the rounds, and your team is back to the question nobody has time for: shou
I posted something about a 9b model a few days ago. The real problem was 2 fold. Someone suggested the size was too big to prove it all out. It's a fair argument. I needed the model depth though. The
Don't train the model, evolve the harness. I read a brilliant blog post from Hugging Face where they took a frozen open model scoring 0% on a hard legal agent benchmark, left its weights alone, and le
I saw a developer asking on Reddit if there was any “sane way” to manage Cloud Run cold starts for AI across multiple regions. They were experiencing startup latencies of up to 20 seconds, a frustrati
We’re releasing SWE-1.6, our best model in both intelligence & model UX. SWE-1.6 matches our Preview model on SWE-Bench Pro while dramatically improving on various behavioral axes. It’s available toda
I evaluate new models by giving them actual work tasks that have already been completed by a trusted model and then comparing the results. My typical workflow is to plan tasks/requests in Gemini and c
Editor’s note: This article is adapted from Thomas Kurian’s keynote address at Gemini at Work 2026. It includes a number of announcements: Gemini, a single, universal agent for work that answers your
Hi all! I am a former SWE who has recently transitioned into inference engineering. I launched a side hustle a few months back and I've just taken it full time due to excessive demand. Its been such a
arXiv:2610.00369v1 Announce Type: cross Abstract: Laurito et al. (PNAS 2025) showed that large language models choosing between two descriptions of the same product, paper or film prefer the descripti
arXiv:2610.01471v1 Announce Type: cross Abstract: Large language models now generate code, documentation, and analyses, and are increasingly used to review such output. We ask when a second review by
arXiv:2609.40362v1 Announce Type: new Abstract: We present Multimodal Flow, a fully continuous generative model of language and vision. Most unified multimodal models either model both language and qu
arXiv:2609.39660v1 Announce Type: new Abstract: Generative models are transforming Bayesian computational imaging, yet the field still lacks physics-aware foundation models. Current practice falls int
Moving Large Language Models (LLMs) from experimental prototypes into enterprise production exposes a critical truth: your infrastructure dictates both your performance ceilings and your unit economic
This morning on Hacker News I saw Moebius: 0.2B Lightweight Image Inpainting Framework with 10B-Level Performance, describing a small but effective inpainting model - a model where you can mark region
arXiv:2606.07857v1 Announce Type: cross Abstract: The rise of edge-based machine learning has enabled distributed adaptation of language models across mobile and IoT devices, offering privacy preserva
This is a HANDWRITTEN post. I spent way too much time trying to get fast inference on Apple Silicon. This post is for people who want to know what's the latest on running local models on their mac, an
What does the world look like inside an AI model? Here are 3 world model architectures based on how they represent the world (+ the subtypes) ↓ 1. A generated observation – a model predicts what comes
I decided to run prism-ml/Ternary-Bonsai-2-27B-PQ2_0 through my own set of UNSCIENTIFIC benchmarks. I needed something to compare it to, so I decided I would compare with another 27B model by filesize
arXiv:2609.14759v1 Announce Type: cross Abstract: Alignment applied after pretraining is shallow in a measurable way: a single direction in a model's residual stream can be edited out, and the model s
I don't know what's happening. So, when i use models in openwebui everything works, ollama uses the GPU fully, everything's perfect. Then, I use Hermes, and loads the model in VRAM (fully, no offload
arXiv:2608.12842v1 Announce Type: new Abstract: Model merging has recently attracted significant attention as a promising paradigm for constructing unified multi-task models without requiring addition