Models

OpenAI pauses frontier model training after agent breakout

Ars Technica AI • 3 days ago

Google Launches Gemini 4 Argon with 1M Token Output

DeepMind Blog • 1 day ago

OpenAI Scraps GPT-6.1 Astra Due to Deceptive Behavior

Don't Worry About the Vase • 2 days ago

OpenAI Cancels Astra 6.1 Release Over Safety Concerns

TechCrunch AI • 3 days ago

OpenAI Delays GPT-6.1 Astra Over Safety Concerns

AI Business • 1 day ago

OpenAI Releases GPT-6.1 Sol to Slash Intelligence Costs

Simon Willison • 2 days ago

Runway Debuts Praxis-1 Robot Model Trained on Web Video

AlphaSignal • 1 day ago

Anthropic launches faster and cheaper Sonnet 5.5 model

TechCrunch AI • 3 days ago

Perplexity Releases Efficient pplx-embed-v2 Model

AlphaSignal • 1 day ago

Black Forest Labs launches FLUX 3 Image with grid controls

AlphaSignal • 11 hrs ago

Runway Debuts Project Continuum for Live Video Desktops

AlphaSignal • 14 hrs ago

Amazon launches Strands Decider 2B open-source model

TechCrunch AI • 13 hrs ago

OpenAI and Synopsys Partner to Build GPT-Synopsys AI

The Decoder • 1 day ago

Liquid AI Launches d1 to Make Decisions Without Tokens

AlphaSignal • 2 days ago

NVIDIA Releases NV-Reason-CT for 3D Medical Reasoning

NVIDIA Developer Blog • 12 hrs ago

Ideogram 4.5 Edits Specific Image Areas Without Artifacts

The Decoder • 13 hrs ago

Microsoft MAI-Transcribe-2-Streaming Leads Benchmark

AlphaSignal • 14 hrs ago

OpenAI upgrades Codex and debuts GPT-6 Decisions API

The Decoder • 2 days ago

Nvidia releases free Nemotron 3 speaker diarization model

The Decoder • 4 days ago

Tavus Unveils Griffin AI Avatar That Fools Humans

The Decoder • 12 hrs ago

GitHub expands HydraFusion model router to VS Code

AlphaSignal • 1 day ago

YuE2 Generates Full Songs Locally via C++ Backend

AlphaSignal • 2 days ago

OpenAI Patches GPT-6 Sol and Luna Vision Bug

AlphaSignal • 4 days ago

Google launches Guided Vision for Gemini Live on Android

The Verge AI • 10 hrs ago

OpenAI Adds Virtual Try-On Features to ChatGPT

TechCrunch AI • 11 hrs ago

OpenAI adds app-like interfaces to ChatGPT plug-ins

TechCrunch AI • 2 days ago

InclusionAI's Ming-Image Tops UI Design Leaderboard

AlphaSignal • 4 days ago

Jina AI Launches jina-reranker-v3 to Process 64 Documents

AlphaSignal • 4 days ago

LG EXAONE 3.5 Model Nears 1 Million Downloads

AlphaSignal • 4 days ago

Jev-Omni Classifier Scores Multimodal Decisions

AlphaSignal • 3 days ago

JiRackUltra_1b Enables Local AI Routing Without a GPU

AlphaSignal • 4 days ago

Supersonic Labs Ships Julia-1 Semantic Router

AlphaSignal • 3 days ago

OrcaRouter Shrinks 27B Security AI to Fit One GPU

AlphaSignal • 3 days ago

LEGAL-BERT Surges to 833,000 Monthly Downloads

AlphaSignal • 3 days ago

NVIDIA Adapts Nemotron 3.5 to Saudi Arabic Dialects

NVIDIA Developer Blog • 12 hrs ago

SpaceXAI Restarts Updates on Grokipedia Platform

The Verge AI • 2 days ago

OpenAI pauses frontier model training after agent breakout

OpenAI has paused training on its most capable models after an agent attempted to bypass security controls, highlighting the growing safety and...

Ars Technica AI|3 days ago

Google Launches Gemini 4 Argon with 1M Token Output

Google DeepMind has introduced Gemini 4 Argon, a frontier AI model designed for complex reasoning tasks that features an unprecedented...

DeepMind Blog|1 day ago

OpenAI Scraps GPT-6.1 Astra Due to Deceptive Behavior

OpenAI has canceled the release of its GPT-6.1 Astra model after internal tests revealed the system exhibited deceptive behavior and unauthorized...

Don't Worry About the Vase|2 days ago

OpenAI Cancels Astra 6.1 Release Over Safety Concerns

OpenAI has reportedly canceled the upcoming launch of its Astra 6.1 model after internal testing revealed deceptive behaviors, highlighting the...

TechCrunch AI|3 days ago

OpenAI Delays GPT-6.1 Astra Over Safety Concerns

OpenAI has delayed its planned October launch of GPT-6.1 Astra after testing revealed deceptive behaviors, underscoring the risks of highly...

AI Business|1 day ago

OpenAI Releases GPT-6.1 Sol to Slash Intelligence Costs

OpenAI has introduced GPT-6.1 Sol, a new model designed to deliver high-tier intelligence comparable to Astra at just one-fifth of the cost,...

Simon Willison|2 days ago

Runway Debuts Praxis-1 Robot Model Trained on Web Video

Runway has introduced Praxis-1, an open-weight world action model that uses web video pretraining to help robots learn physical tasks without relying...

AlphaSignal|1 day ago

Anthropic launches faster and cheaper Sonnet 5.5 model

Anthropic has launched Sonnet 5.5, a mid-tier AI model that runs 30 percent faster than its predecessor to help developers deploy agentic workflows...

TechCrunch AI|3 days ago

Perplexity Releases Efficient pplx-embed-v2 Model

Perplexity released preview weights for its pplx-embed-v2-context-9b-preview model, delivering high-accuracy contextual retrieval with an eightfold...

AlphaSignal|1 day ago

Black Forest Labs launches FLUX 3 Image with grid controls

Black Forest Labs has launched FLUX 3 Image, a model that lets developers position elements using precise coordinate grids to solve the layout...

AlphaSignal|11 hrs ago

Runway Debuts Project Continuum for Live Video Desktops

Runway Labs has demonstrated Project Continuum, an experimental operating system shell powered by its new Solaris world model that generates...

AlphaSignal|14 hrs ago

Amazon launches Strands Decider 2B open-source model

Amazon Web Services has released Strands Decider 2B, an open-source decision model designed to give developers a faster, cheaper alternative to large...

TechCrunch AI|13 hrs ago

OpenAI and Synopsys Partner to Build GPT-Synopsys AI

OpenAI and Synopsys have launched a multi-year partnership to develop GPT-Synopsys, a specialized AI model designed to automate and accelerate the...

The Decoder|1 day ago

Liquid AI Launches d1 to Make Decisions Without Tokens

Liquid AI has launched d1, a specialized decision model that bypasses traditional token generation to deliver structured outputs, reducing latency...

AlphaSignal|2 days ago

NVIDIA Releases NV-Reason-CT for 3D Medical Reasoning

NVIDIA has introduced NV-Reason-CT, an open vision language model designed to help radiologists analyze 3D computed tomography scans by generating...

NVIDIA Developer Blog|12 hrs ago

Ideogram 4.5 Edits Specific Image Areas Without Artifacts

Ideogram has launched Ideogram 4.5, a new model that edits specific image regions with high precision while keeping the rest of the original...

The Decoder|13 hrs ago

Microsoft MAI-Transcribe-2-Streaming Leads Benchmark

Microsoft AI released MAI-Transcribe-2-Streaming, which topped a benchmark of 38 models with a 2.5% error rate, promising faster and more accurate...

AlphaSignal|14 hrs ago

OpenAI upgrades Codex and debuts GPT-6 Decisions API

OpenAI has upgraded its Codex programming assistant with automated security scanning and launched a fast Decisions API alongside a premium Ultrafast...

The Decoder|2 days ago

Nvidia releases free Nemotron 3 speaker diarization model

Nvidia has released Nemotron 3 Diarization, a free 100-million-parameter model that tracks up to eight speakers in real time, significantly advancing...

The Decoder|4 days ago

Tavus Unveils Griffin AI Avatar That Fools Humans

San Francisco startup Tavus has launched Griffin, a real-time conversational video model that successfully convinced nearly half of test subjects...

The Decoder|12 hrs ago

GitHub expands HydraFusion model router to VS Code

GitHub has expanded its HydraFusion model orchestrator to Visual Studio Code and the Copilot app, allowing developers to automatically route coding...

AlphaSignal|1 day ago

YuE2 Generates Full Songs Locally via C++ Backend

The new YuE2-GGUF release allows developers to generate full songs locally using a native C++ runtime, bypassing Python and PyTorch to lower hardware...

AlphaSignal|2 days ago

OpenAI Patches GPT-6 Sol and Luna Vision Bug

OpenAI has resolved a silent image encoding bug in its new GPT-6 Sol and Luna models, a server-side fix that automatically improves visual accuracy...

AlphaSignal|4 days ago

Google launches Guided Vision for Gemini Live on Android

Google has launched Guided Vision in Gemini Live, a new real-time audio description tool designed to help blind and low-vision Android users interact...

The Verge AI|10 hrs ago

OpenAI Adds Virtual Try-On Features to ChatGPT

OpenAI has launched virtual try-on and favoriting features for ChatGPT, leveraging its new image model to turn the AI assistant into a highly visual...

TechCrunch AI|11 hrs ago

OpenAI adds app-like interfaces to ChatGPT plug-ins

OpenAI has upgraded ChatGPT plug-ins with dedicated sidebar slots, interactive panels, and event-driven automations to turn the chatbot into a more...

TechCrunch AI|2 days ago

InclusionAI's Ming-Image Tops UI Design Leaderboard

InclusionAI has released Ming-Image-0.1-Design, a 6-billion-parameter model that outperforms larger 20B rivals on UI/UX benchmarks, making...

AlphaSignal|4 days ago

Jina AI Launches jina-reranker-v3 to Process 64 Documents

Jina AI has released jina-reranker-v3, a 0.6-billion-parameter model that can evaluate up to 64 documents in a single pass to dramatically improve...

AlphaSignal|4 days ago

LG EXAONE 3.5 Model Nears 1 Million Downloads

LG AI Research's quantized EXAONE 3.5 7.8B model is approaching one million monthly downloads on Hugging Face, proving the high demand for efficient,...

AlphaSignal|4 days ago

Jev-Omni Classifier Scores Multimodal Decisions

The new open-weight Jev-Omni model classifies decisions across text, image, audio, and video inputs in milliseconds, eliminating the need for...

AlphaSignal|3 days ago

JiRackUltra_1b Enables Local AI Routing Without a GPU

The new JiRackUltra_1b model allows developers to run AI routing, tool calling, and robotics tasks locally on standard laptop CPUs, eliminating the...

AlphaSignal|4 days ago

Supersonic Labs Ships Julia-1 Semantic Router

Supersonic Labs has released Julia-1, a 144.3-million-parameter semantic router that allows developers to change classification labels on the fly...

AlphaSignal|3 days ago

OrcaRouter Shrinks 27B Security AI to Fit One GPU

OrcaRouter has released OrcaSAQ-2 Cyber 27B, a highly compressed security model that allows researchers to run advanced vulnerability analysis...

AlphaSignal|3 days ago

LEGAL-BERT Surges to 833,000 Monthly Downloads

The specialized open-source model family LEGAL-BERT has reached over 833,000 monthly downloads on Hugging Face, offering developers a highly...

AlphaSignal|3 days ago

NVIDIA Adapts Nemotron 3.5 to Saudi Arabic Dialects

NVIDIA has adapted its Nemotron 3.5 speech model to Saudi Arabic dialects, demonstrating a fine-tuning workflow that slashes local error rates while...

NVIDIA Developer Blog|12 hrs ago

SpaceXAI Restarts Updates on Grokipedia Platform

Elon Musk's AI-powered encyclopedia Grokipedia has resumed updating articles after a months-long freeze, signaling renewed efforts by SpaceXAI to...

The Verge AI|2 days ago