Skip to content
gentic.news — AI News Intelligence Platform
Connecting to the Living Graph…

AI News Digest

Thursday, April 23, 2026

18 stories covered by gentic.news intelligence

A Tencent employee demonstrates the HY3 AI model interface on a large screen at a tech conference, showing Yuanbao…
Big Tech
100

Tencent's HY3 AI Model Has 295B Params, Led by Ex-OpenAI Researcher

Tencent unveiled its HY3 preview model, its most powerful yet with 295 billion parameters. It's already deployed in consumer app Yuanbao and coding assistant CodeBuddy.

scmp.com/Apr 23, 2026/3 min read/Widely Reported
model releaseleadershipbusiness ai
A developer's hands typing on a laptop keyboard with a code editor on screen, next to the Anthropic logo and a…
Products & Launches
100

Anthropic Removes Claude Code from $20 Plan, Signals AI Pricing Shift

Anthropic removed its AI coding tool Claude Code from the $20/month Pro plan, moving it to $100+ tiers. This reflects the high operational costs of AI coding assistants and signals a broader industry pricing shift.

pub.towardsai.net/Apr 23, 2026/3 min read/Widely Reported
claude codeanthropicagentic coding
A sleek metallic robot hand reaches toward a glowing blue digital brain floating above a futuristic city skyline at dusk
Products & Launches
97

OpenAI Launches GPT-5.5: Smarter Agents, Deeper Tool Use

OpenAI unveiled GPT-5.5, positioned as a new intelligence tier designed for real-world work and autonomous agents, with enhanced tool-use capabilities and complex goal understanding.

x.com/Apr 23, 2026/3 min read
gpt-5.5model releaseai agents
MIT researcher presents Recursive Language Model diagram on screen, showing code-based data search and filtering…
AI Research
95

MIT's RLM Handles 10M+ Tokens, Outperforms RAG on Long-Context Benchmarks

MIT researchers introduced Recursive Language Models (RLMs), which treat long documents as an external environment and use code to search, slice, and filter data, achieving 58.00 on a hard long-context benchmark versus 0.04 for standard models.

x.com/Apr 23, 2026/3 min read
researchai architecturemit
Researchers demonstrating Meta's Sapiens2 vision transformers on a monitor, showing human pose estimation and…
AI Research
92

Meta's Sapiens2: 1B Human Image ViTs for Pose, Segmentation, Normals

Meta open-sourced Sapiens2 on Hugging Face, a family of vision transformers pretrained on 1 billion human images for pose estimation, segmentation, normal estimation, and point maps. The models target high-resolution human-centric perception.

x.com/Apr 23, 2026/3 min read
open sourcevision transformercomputer vision
A child smiling at a laptop screen showing Claude AI tutoring interface, with Khan Academy lesson materials visible…
Products & Launches
89

Claude Now Tutors Kids for Free, Matching $100/hr Private Lessons

Claude can now teach kids any school subject like a $100/hour private tutor from Khan Academy, for free. This brings high-quality, personalized AI tutoring to anyone with internet access.

x.com/Apr 23, 2026/3 min read/Multi-Source
claudeanthropicconsumer ai
A futuristic digital brain made of glowing blue neural nodes and circuits, floating above a keyboard with code on a…
Products & Launches
87

OpenAI Teases GPT-5.5 Launch: What We Know

A tweet from @intheworldofai suggests OpenAI will launch GPT-5.5 tomorrow, framing it as a pivotal moment akin to GPT-3.5. The announcement signals a significant model upgrade, though details remain scarce.

x.com/Apr 23, 2026/3 min read
product launchai modelsgpt
Diagram comparing stateful memory with event-sourced stateless architecture, showing multiple AI agents connecting…
AI Research
85

Stateless Memory for Enterprise AI Agents: Scaling Without State

The paper replaces stateful agent memory with immutable decision logs using event-sourcing, allowing thousands of concurrent agent instances to scale horizontally without state bottlenecks.

x.com/Apr 23, 2026/3 min read
scalabilityai agentssystem design
Two corporate office towers with Meta and Microsoft logos, employees carrying boxes exit as a giant robot arm…
Products & Launches
85

Meta, Microsoft Lay Off 17,000 in One Day for AI Spending

Meta fired 8,000 employees and Microsoft laid off 9,000 within hours of each other, signaling a coordinated shift of resources from headcount to AI compute and model development. The layoffs underscore a trend where big tech prioritizes AI investment over workforce stability.

x.com/Apr 23, 2026/3 min read
layoffsmetamicrosoft
Cua Driver open-source macOS agent controls apps via visual UI, shown in a screenshot of an app management interface…
Products & Launches
85

Cua Driver Open-Sourced: macOS Agent Control for Any App

Cua released Cua Driver as open-source, allowing agents like Claude Code and Codex to drive any macOS app through visual understanding and direct UI interaction.

x.com/Apr 23, 2026/3 min read
open sourceautomationai agents
A person holds a smartphone displaying a chat interface with an AI assistant, while a stylized potato icon floats…
AI Research
85

GPT-5.5 'Spud' Prioritizes Pretraining Over Chain-of-Thought

A new OpenAI model, Spud (GPT-5.5), focuses on pretraining improvements rather than heavy test-time compute, promising faster and cheaper responses.

x.com/Apr 23, 2026/3 min read
efficiencyaimodel training
A person handing a document to a sleek, humanoid robot in a modern office setting, with the robot's hand raised to…
Products & Launches
85

Delegate Launches: An AI Agent You Hand Work To and Walk Away

A new AI agent called Delegate lets users assign work and walk away, with the agent handling execution autonomously. The launch signals a shift toward hands-off AI assistants that manage complex tasks independently.

x.com/Apr 23, 2026/3 min read
launchproductai agents
A humanoid robot in a lab environment performs a dynamic motion sequence, with motion lines and 3D skeletal overlays…
AI Research
85

NVIDIA Open-Sources Motion Diffusion Model for Humanoid Robots

NVIDIA open-sourced Kimono, a motion diffusion model for humanoid robots, trained on 700 hours of motion capture data. It generates 3D human and robot motions from text prompts, supports keyframe and end-effector control, and runs on Unitree G1.

x.com/Apr 23, 2026/3 min read
open sourcemotion generationrobotics
A sleek dark interface displays Claude Opus 4.7 benchmark scores, with a bar chart showing 87.6% on SWE-Bench…
Products & Launches
84

Anthropic Opus 4.7: 87.6% SWE-Bench, Constrained Cyber Capabilities

Anthropic released Claude Opus 4.7 on April 16, 2026, achieving 87.6% on SWE-Bench Verified and 64.3% on SWE-Bench Pro — leading GPT-5.4 and Gemini 3.1 Pro. The company also confirmed it deliberately constrained cybersecurity capabilities in Opus 4.7, with the more powerful Mythos Preview model (83.1% on CyberGym) restricted to select partners.

pub.towardsai.net/Apr 23, 2026/3 min read
claudecodinganthropic
Horizon Robotics CEO presenting a chip, open-source OS, and smart driving system at a tech launch event, with a…
Products & Launches
82

Horizon Launches Full-Stack AI Platform for Autonomous Driving

Horizon Robotics launched a trio of products—a new chip, an open-source OS, and a smart driving system—aiming to push cars closer to becoming autonomous AI agents. The platform integrates hardware and software for enhanced perception and decision-making.

pandaily.com/Apr 23, 2026/3 min read/Multi-Source
open sourceautonomous drivingchips
A bar chart showing AI model accuracy scores on thermodynamic reasoning, with Claude Opus leading at 94.1% and other…
AI Research
78

ThermoQA Benchmark Reveals LLM Reasoning Gaps: Claude Opus Leads at 94.1%

Researchers released ThermoQA, a 293-question benchmark testing thermodynamic reasoning. Claude Opus 4.6 scored 94.1% overall, but models showed significant degradation on complex cycle analysis versus simple property lookups.

arxiv.org/Apr 23, 2026/3 min read/Multi-Source
large-language-modelsanthropicresearch
A network of glowing blue nodes and edges on a dark background, representing a graph database structure powering…
Products & Launches
77

FalkorDB: Graph Database for Multi-Hop AI Queries in Milliseconds

FalkorDB, an open-source graph database, stores connections as a sparse matrix to accelerate multi-hop queries by 100x. Combined with built-in vector search, it enables GraphRAG systems that answer complex relational questions without pre-built articles.

x.com/Apr 23, 2026/3 min read
open-sourcevector-searchdatabases
A smartphone screen displays a text conversation with Cappy, showing video editing instructions sent via SMS and a…
Products & Launches
75

Mirage's Cappy Edits Video via Text Message with No App

Mirage launched Cappy, a text-based video editing service that delivers fully edited videos via SMS. This first-of-its-kind approach eliminates traditional editing interfaces entirely.

x.com/Apr 23, 2026/3 min read
ai videoediting toolsconsumer ai

Recent Daily Digests