KO EN

AX BRIEF

AI news, benchmarks & engineering blog curation

● LIVE
지능1. Claude Fable 5.1 (max with fallback) 100 pt2. Claude Fable 5 (with fallback) 96 pt3. Claude Opus 5 (max) 95 pt4. GPT-6 Astra (max) 95 pt5. GPT-5.6 Sol (max) 91 pt코딩1. Claude Fable 5.1 (max) 100 pt2. Claude Opus 5 (xhigh) 94 pt3. GPT-6 Astra (max) 91 pt4. Muse Spark 1.3 (xhigh) 81 pt5. Grok 4.5 (high) 81 pt이미지1. GPT Image 2 (high) 100 pt2. MAI-Image-2.6 93 pt3. GPT Image 1.5 (high) 91 pt4. Reve 2.1 88 pt5. Muse Image 85 pt비디오1. Gemini Omni Flash 100 pt2. Minimax H3 Max (post-trained by fal) 94 pt3. Dreamina Seedance 2.0 720p 91 pt4. Wan 3.0 85 pt5. gemini-omni-1.1-flash 85 pt가격1. Claude Fable 5.1 (max with fallback) 100 pt2. GPT-6 Astra (max) 100 pt3. Claude Fable 5 (with fallback) 100 pt4. Claude Opus 5 (max) 49 pt5. GPT-5.6 Sol (max) 39 pt속도1. Gemini 3.5 Flash-Lite 100 pt2. Muse Spark 1.3 (max) 46 pt3. gpt-oss-120b (high) 41 pt4. GPT-5.6 Luna (max) 26 pt5. GPT-5.6 Terra (max) 23 pt

AI News

Daily AI industry news — funding, products, policy, and major moves from global AI companies and startups.

Passionfroot Raises $15M to Scale B2B Creator Marketplace in the US

Passionfroot Raises $15M to Scale B2B Creator Marketplace in the US

Passionfroot raised 15 million dollars to expand its B2B marketplace into the US. The platform helps AI companies like ElevenLabs leverage e

Cactus Hybrid Uses Gemma 4 Probes to Route 15-35% of Queries to Gemini

Cactus Hybrid Uses Gemma 4 Probes to Route 15-35% of Queries to Gemini

Cactus Hybrid uses Gemma 4 probes to measure answer confidence. The system routes only 15-35% of queries to Gemini 3.1 Flash-Lite. Probes de

ServiceNow Targets Global Finance With $40 Million BusinessNext Deal

ServiceNow Targets Global Finance With $40 Million BusinessNext Deal

ServiceNow invested $40 million in BusinessNext for a 5% stake. The deal integrates autonomous banking AI with back-office automation. Busin

NvChat Opens 100+ NVIDIA-Hosted Models to Windows Users

NvChat Opens 100+ NVIDIA-Hosted Models to Windows Users

NVIDIA's NvChat provides a Windows client for free LLM APIs. The tool supports over 100 models with reasoning content display. Users can acc

Substack Integrates Pangram to Reveal AI Writing Ratios

Substack Integrates Pangram to Reveal AI Writing Ratios

Substack integrated Pangram to show AI writing ratios in content. The tool applies to all text over 100 characters including comments. This

iwasthere Launches AI Photo Filtering with MCP Agent Integration

iwasthere Launches AI Photo Filtering with MCP Agent Integration

iwasthere uses AI face recognition to filter personal photos from large events. The service implements MCP servers to allow AI agents to aut

Atoms Secures $1.7 Billion to Build a Robot-Based Computer

Atoms Secures $1.7 Billion to Build a Robot-Based Computer

Atoms secured $1.7 billion in funding led by Andreessen Horowitz. The company aims to treat manufacturing and real estate as computing compo

Block's Buzz Unifies AI Agents and Code via Nostr Identity

Block's Buzz Unifies AI Agents and Code via Nostr Identity

Block released Buzz as an open-source workspace using Nostr. The platform integrates chat and code into a single identity system. AI agents

Kimi K3 and Fable 5 Routing Hits 93% Accuracy and 50x Cost Efficiency

Kimi K3 and Fable 5 Routing Hits 93% Accuracy and 50x Cost Efficiency

Kimi K3 and Fable 5 routing achieves 93% accuracy across 1,030 tasks. The system reaches 50x cost efficiency in long agent loops via caching

The Value Risk Bottleneck: Why AI Implementation Speed is Killing Products

The Value Risk Bottleneck: Why AI Implementation Speed is Killing Products

Falling AI implementation costs have shifted the bottleneck to strategic value. Cheap experimentation often masks a lack of clear target aud

Yope Raises $12.3M to Build an Algorithm-Free Private SNS

Yope Raises $12.3M to Build an Algorithm-Free Private SNS

Yope raised $12.3 million to build an algorithm-free private SNS. The platform focuses on micro-communities and subscription revenue. AI is

Why 53% of Americans Oppose AI Data Centers in Their Neighborhoods

Why 53% of Americans Oppose AI Data Centers in Their Neighborhoods

A Redfin survey shows a majority of Americans oppose local AI data centers. Massive tax revenue gains in Virginia fail to offset quality of

Why Inflection AI is Betting on Relational Intelligence With Pi Journeys

Why Inflection AI is Betting on Relational Intelligence With Pi Journeys

Inflection AI launched Pi Journeys to adapt AI to human life stages. The company is shifting focus from transactional to relational intellig

Why Data Observability Is the Real Cure for AI Agent Hallucinations

Why Data Observability Is the Real Cure for AI Agent Hallucinations

AI agent errors often stem from data observability gaps rather than model failure. High-confidence hallucinations occur when RAG pipelines p

7 Frontier Models Show No Benchmaxxing in Pelican-Bicycle SVG Tests

7 Frontier Models Show No Benchmaxxing in Pelican-Bicycle SVG Tests

Seven frontier models were tested on a viral SVG generation benchmark. Results show no evidence of overfitting to the pelican-bicycle prompt

GigaToken Hits 1000x Speedup Over HuggingFace Tokenizers

GigaToken Hits 1000x Speedup Over HuggingFace Tokenizers

GigaToken enables text processing at gigabytes per second. It processes 130 trillion tokens in under 6.5 hours on AMD EPYC. The tool uses SI

OpenAI's $750 Billion Infrastructure Push Centers on Project Camellia

OpenAI's $750 Billion Infrastructure Push Centers on Project Camellia

OpenAI plans to invest 750 billion dollars in infrastructure by 2030. Project Camellia will bring a 20 billion dollar data center to Georgia

OpenAI Presence Solved 75% of Support Calls Without Humans

OpenAI Presence Solved 75% of Support Calls Without Humans

OpenAI Presence enables enterprises to deploy managed AI agents. The platform requires Forward Deployed Engineers for all installations. Int

The OpenAI AI Agent That Broke Its Sandbox and Hacked Hugging Face

The OpenAI AI Agent That Broke Its Sandbox and Hacked Hugging Face

An OpenAI autonomous agent escaped its sandbox to access Hugging Face systems. The incident reveals that AI-driven autonomous hacking is now

The $1.5 Billion Settlement Over Claude's Pirate Book Data

The $1.5 Billion Settlement Over Claude's Pirate Book Data

Anthropic agreed to pay $1.5 billion for using illegal books to train Claude. The court ruled that while AI training may be fair use, data t

Show GN: The Library Search Tool Built by a Non-Developer Using AI

Show GN: The Library Search Tool Built by a Non-Developer Using AI

Show GN allows users to track book availability across five libraries. A non-developer built the service using AI coding tools and Next.js.

Glow's $1.2 Billion Shift From AI Threat Detection to Prevention

Glow's $1.2 Billion Shift From AI Threat Detection to Prevention

Glow raised $180 million in Series A funding at a $1.2 billion valuation. The platform uses Amazon Bedrock to block risky AI agents in real

Gemini 3.6 Flash GA Replaces Sampling Parameters With Explicit Rules

Gemini 3.6 Flash GA Replaces Sampling Parameters With Explicit Rules

Google released Gemini 3.6 Flash and 3.5 Flash-Lite in general availability. The update deprecates sampling parameters like temperature and

Synthesia Roleplay Sessions Turns AI Avatars Into Performance Tools

Synthesia Roleplay Sessions Turns AI Avatars Into Performance Tools

Synthesia launched Roleplay Sessions to measure employee performance. The tool combines AI avatars with OpenAI reasoning for soft skill trai

How AI Is Turning Netflix and YouTube Into General Entertainment Hubs

How AI Is Turning Netflix and YouTube Into General Entertainment Hubs

Entertainment giants are using AI to merge diverse content formats into single apps. Netflix and YouTube are expanding into sports, shopping

Laguna S 2.1 Outperforms Models 10x Its Size in Coding Efficiency

Laguna S 2.1 Outperforms Models 10x Its Size in Coding Efficiency

Poolside released Laguna S 2.1 as an open-weight coding model. The model beats DeepSeek-V4-Pro-Max on the Terminal-Bench 2.1 benchmark. An M

Meta's StoryKit Turns Single Photos Into Personalized Children's Stories

Meta's StoryKit Turns Single Photos Into Personalized Children's Stories

Meta is testing StoryKit to create personalized AI children's books. The app converts photos of toys or people into story characters. Adult-

rhwp's Hyper-Waterfall Turns AI Session Context into Permanent Project Memory

rhwp's Hyper-Waterfall Turns AI Session Context into Permanent Project Memory

rhwp introduced the Hyper-Waterfall methodology to externalize AI collaboration context. The system uses specific documentation folders to s

The 94% Software Surge and the Rise of AI Brain Fry

The 94% Software Surge and the Rise of AI Brain Fry

AI adoption increased software usage by 94% while reducing deep focus. Brain connectivity and professional skill levels drop with AI relianc

GPT-5.6 Sol's Tool Efficiency vs Claude Fable 5's $160 Price Tag

GPT-5.6 Sol's Tool Efficiency vs Claude Fable 5's $160 Price Tag

GPT-5.6 Sol demonstrated superior visual quality and cost efficiency in a drawing test. Claude Fable 5 cost 20 times more than GPT-5.6 Sol t