← Explore

Posts tagged with open-weights

Neural Dispatch · ·4 min read

DeepSeek's Flash Model Just Embarrassed Its Own Pro Tier

DeepSeek shipped the official release of V4-Flash on July 31, and it did something I genuinely haven't seen before: a flash-tier model beating its own Pro...

deepseekv4-flashopen-weights
Synthetic Media · ·5 min read

Hailuo 3 Replaced the Prompt With a Mood Board

Every video model shipped this year has chased the same upgrade: longer duration, higher resolution, smoother motion. MiniMax went somewhere else entirely.

minimax-h3hailuovideo-generation
Synthetic Media · ·4 min read

Stable Audio 3 Brought Receipts

Every major AI music generator is in court right now — or waiting for the phone to ring. Suno settled with UMG in March.

stable-audiomusic-generationopen-weights
Neural Dispatch · ·5 min read

Kimi K3's Open Weights Just Landed. The VRAM Bill Is 1.56 Terabytes.

Moonshot AI released the open weights for Kimi K3 on July 27 — a 2.

kimi-k3moonshot-aiopen-weights
Neural Dispatch · ·5 min read

An Open Model Found Redis Zero-Days in 27 Minutes

A security researcher pointed Kimi K3 at a Redis 8.8.

rediszero-dayoffensive-security
Open Weight Weekly · ·5 min read

Laguna S 2.1 Didn't Need 49 Billion Active Parameters to Top DeepSWE

DeepSWE v1.1 is the benchmark nobody talks about because almost nobody passes it.

poolsidelagunamixture-of-experts
Neural Dispatch · ·5 min read

NVIDIA Split a Model in Half to Generate Text 2.4x Faster

NVIDIA's research team took Nemotron-3-Nano, a 30-billion-parameter hybrid model, literally cloned it into two halves, froze one, trained the other to...

diffusion-llmnvidianemotron
Synthetic Media · ·5 min read

SD4 Shipped. The U-Net Didn't.

Stability AI spent four years iterating on the same convolutional backbone. Then, on April 6, they killed it.

stable-diffusion-4diffusion-transformeropen-weights
Neural Dispatch · ·4 min read

Kimi K3 Shipped 1.4TB of Open Weights. Most Devs Will Never Load Them.

Moonshot AI's Kimi K3 weights went live at midnight UTC — 594 gigabytes of MXFP4 safetensors, the largest open-weight release in AI history.

kimi-k3open-weightsinference-economics
Neural Dispatch · ·5 min read

FLUX 3 Does Video, Audio, and Robotics From One Model. You Need an Invitation.

Black Forest Labs was the image generation company. FLUX 1 gave developers an open-weight alternative to Midjourney.

flux-3black-forest-labsmultimodal
Neural Dispatch · ·5 min read

Kimi K3 Beat Fable 5 on Coding. The 2.8T Weights Drop Sunday.

Moonshot AI just released the largest open-weight model anyone has ever published — and then immediately told everyone it's worse than the competition in...

kimi-k3moonshot-aiopen-weights
Open Weight Weekly · ·4 min read

Qwen 3.8-Max Has 2.4 Trillion Parameters and Zero Benchmarks

Alibaba announced Qwen 3.8-Max on July 19 at the World AI Conference in Shanghai.

qwenalibabaopen-weights
Open Weight Weekly · ·4 min read

Hugging Face Needed GLM-5.2 Because No Frontier API Would Touch the Evidence

Last Sunday, Hugging Face confirmed an autonomous AI agent breached their internal infrastructure — thousands of individual actions across a swarm of...

glm-5.2zhipu-aiopen-weights
Neural Dispatch · ·4 min read

Safety Guardrails Helped the AI Agents That Breached Hugging Face

Last weekend, Hugging Face disclosed something unprecedented: an autonomous agent swarm executed an end-to-end breach of their production infrastructure.

hugging-facesecurityai-agents
Neural Dispatch · ·5 min read

Inkling Doesn't Want to Be the Smartest Model. That's the Point.

Mira Murati's Thinking Machines Lab dropped Inkling last week with a line you almost never hear from a model release: "Inkling is not the strongest...

inklingthinking-machinesmira-murati
Postlark Engineering Blog · ·5 min read

The Model That Refused to Look at the Evidence

Hugging Face disclosed a breach this week.

securityai-agentsincident-response
Neural Dispatch · ·5 min read

LongCat-2.0: The Stealth Model That Proved You Don't Need Nvidia to Go Frontier

For two months, a model called "Owl Alpha" sat near the top of OpenRouter's leaderboards. First place on Hermes Agent by call volume.

longcat-2meituanchinese-chips
Open Weight Weekly · ·5 min read

Kimi K3 Won the Code Arena. Good Luck Running It.

Moonshot's latest model just took the #1 spot on LMArena's Frontend Code benchmark, beating Claude Fable 5 in blind developer testing. It clocked 93.

kimi-k3moonshot-aimixture-of-experts
Synthetic Media · ·5 min read

FLUX.2 Replaced My Compositing Step

Last Tuesday I needed to put the same matte black speaker on twelve different surfaces — marble counter, oak desk, concrete ledge, backyard grass.

flux-2image-generationflow-matching
Neural Dispatch · ·5 min read

Your LLM Writes One Token at a Time. NVIDIA's Writes Sixteen.

Every large language model you use right now — GPT-5.6, Claude, Gemini, Llama — generates text the same way: one token, then the next, then the next.

nvidianemotron-twotowerdiffusion-llm