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

Latest AI News

Every story, newest first — research, funding, product launches, policy, and analysis.

Close-up of a Final Cut Pro timeline on a computer screen, showing multiple video tracks with colored clips and a…
AI Research
65

AI editor matches pro on 84% of video cuts in blind test

AI editor matched pro on 84% of video cuts in blind test of 4-hour project. Suggests editorial judgment is partially automatable.

x.com/6h ago/3 min read
ai videovideo editingbenchmark
Side-by-side code editor windows show three AI agent implementations: Cursor with 9 lines, Claude Code with 47…
Opinion & Analysis
72

9-Line Agent: Cursor Beats Claude, OpenAI SDKs in Dev Build Test

A developer built the same agent in Cursor (9 lines), Claude Code (47 lines), and OpenAI Codex (31 lines). The gap is in tool orchestration architecture, not model capability.

pub.towardsai.net/13h ago/3 min read
developer experienceproduct architectureai coding tools
Smartphone displaying LLaDA-8B inference interface with latency reduction metrics, NPU chip schematic overlay
AI Research
80

llada.cpp Cuts LLaDA-8B Latency 17-42x on Mobile NPU

llada.cpp, the first NPU-aware dLLM inference framework, cuts LLaDA-8B latency 17-42x on smartphones, enabling real-time on-device generation.

arxiv.org/16h ago/3 min read/Multi-Source
ai inferencemobile hardwarediffusion models
A sleek robotic arm on a lab bench precisely assembles a small electronic circuit board, with glowing blue lights…
AI Research
72

WorkBench Revisited: Claude Opus 4.8 Hits 89% Task Completion

Claude Opus 4.8 completes 89% of WorkBench tasks with 2.5% harm rate, up from GPT-4's 43% and 26% in 2024, showing capability and safety align.

arxiv.org/16h ago/3 min read
anthropicagent safetybenchmarks
Mirage Probes Paper Reveals Two Distinct VLM Failure Modes
AI Research
82

Mirage Probes Paper Reveals Two Distinct VLM Failure Modes

Mirage Probes paper reveals VLMs have two distinct failure modes—textual biases and spurious images—requiring different mitigations. Text cleaning only fixes one; the other needs representational interventions.

arxiv.org/16h ago/3 min read/Multi-Source
ai safetycomputer visionresearch
Bar chart comparing AI model scores on MA-ProofBench, with GPT-5.5 reaching 16% on undergraduate and 5% on PhD…
AI Research
72

MA-ProofBench: GPT-5.5 Hits 16% on Math Analysis, Most Models Near 0%

MA-ProofBench, a new theorem-proving benchmark for mathematical analysis, shows GPT-5.5 achieving 16% on undergraduate problems and 5% on PhD-level, with most models near 0% on the harder set.

arxiv.org/16h ago/3 min read
mathematicstheorem provingbenchmarks
Microsoft CEO Satya Nadella gestures while speaking in an interview, with a blurred background of computer screens…
Opinion & Analysis
80

Nadella: AI's New Unit Is 'Tokens per Dollar per Watt'

Satya Nadella defined AI's supply-side economics as 'Tokens per Dollar per Watt', urging infrastructure focus for companies, industries, and countries.

x.com/23h ago/3 min read
energyinfrastructuremicrosoft
A line graph titled SWE-Explore showing low coverage rates around 14-19% for Claude Code, Codex 5.3, and OpenHands…
AI Research
92

SWE-Explore: AI coding agents find files but miss 81-86% of critical lines

SWE-Explore benchmark shows Claude Code, Codex cover only 14-19% of critical lines despite finding the right file. Model strength doesn't fix the structural weakness.

the-decoder.com/1d ago/3 min read/Widely Reported
code generationresearchai agents
A laptop screen displays code from Zhipu AI's GLM-5.2 model, with a diagram of a 1M token context window and an MIT…
Open Source
98

Zhipu AI Open-Sources GLM-5.2 with 1M Token Context Under MIT License

Zhipu AI open-sourced GLM-5.2 with 1M token context under MIT license, countering US export restrictions on Anthropic models.

pandaily.com/1d ago/3 min read/Widely Reported
open-sourceanthropiczhipu ai
A futuristic digital rendering of a robotic arm manipulating glowing tokens on a simulated grid, with Nvidia…
AI Research
87

Nvidia Cosmos 3 Unifies Physical AI — Action as Token

Nvidia's Cosmos 3 unifies physical AI perception, simulation, and action in one model via action-as-token. No benchmark data disclosed yet.

x.com/1d ago/3 min read
physical-airoboticsworld-models
A Huawei executive on a conference stage gestures toward a large screen displaying the HarmonyOS 7 logo and a…
Products & Launches
94

Huawei HarmonyOS 7 Ships 2,100 System-Level AI Agent Capabilities

Huawei launched HarmonyOS 7 with Xiaoyi as a system-level AI agent exposing 2,100 capabilities, shifting from app-centric to intent-driven interaction.

pandaily.com/1d ago/3 min read/Widely Reported
operating systemsai agentsmobile
Infographic illustrating OpenRouter's Fusion API routing queries across multiple AI providers to achieve high…
Products & Launches
87

OpenRouter Fusion API Claims Fable-Level IQ at Half the Cost

OpenRouter's Fusion API routes queries across providers to match Fable-level intelligence at half the cost, per company claims. No third-party benchmarks disclosed.

x.com/1d ago/3 min read
ai infrastructurecost optimizationmodel routing
A person typing on a laptop with code on the screen, likely representing the AI model Rio 3.5 Open 397B from Rio de…
AI Research
85

Rio 3.5 Open 397B: City Gov't Model Tops Open-Source Leaderboard

Rio 3.5 Open 397B claims SOTA open-source status but lacks any evidence. A single tweet is the only source.

x.com/1d ago/3 min read
municipal-aiopen-source-llmbenchmark-claims
AI researcher Rohan Paul posts on X that US government ordered Anthropic to shut down its strongest Claude models…
Products & Launches
99

US Gov’t Orders Anthropic to Shut Down Strongest Claude Models

US ordered Anthropic to shut down strongest Claude models via export controls. No official confirmation yet.

x.com/2d ago/3 min read/Multi-Source
anthropicfrontier modelsexport controls
Bar chart comparing BIRD benchmark scores shows Gemini-SQL2 at 80.04% and GPT-5.5-xhigh around 73%, with Gemini 3.1…
AI Research
95

Google Gemini-SQL2 Hits 80.04% on BIRD, Beating GPT-5.5 by 7 Points

Google's Gemini-SQL2 hits 80.04% on BIRD, beating GPT-5.5 by 7 points and Claude Opus 4.6 by 9 points, with no public release or paper yet.

the-decoder.com/2d ago/3 min read/Widely Reported
nl-to-sqlfoundation-modelsbenchmarks
Nvidia executive at podium denies chip smuggling claims, with map of Latin America and China on screen behind
Policy & Ethics
95

Nvidia Denies Anthropic's China Chip Smuggling Claims via Latin America

Nvidia's Latin America chief denied Anthropic's allegations of chip smuggling to China via the region, expressing frustration with U.S. export controls. The denial highlights tensions between AI safety and hardware sales.

pandaily.com/2d ago/3 min read/Widely Reported
geopoliticsai hardwareregulation
A sleek credit card designed with AI motifs, adorned with the Kimi chatbot logo, held by a hand against a futuristic…
Products & Launches
90

Moonshot AI, State Bank Launch First AI-Native Credit Card in China

Moonshot AI's Kimi launches world's first AI-native credit card with state-owned bank, converting spending into compute credits.

pandaily.com/2d ago/3 min read/Widely Reported
ai monetizationconsumer aichina ai
A doctor in a white coat studies a tablet displaying AI diagnostic data, surrounded by medical charts and a…
AI Research
81

General LLMs Beat Clinical AI Tools in Doctor Study

Frontier LLMs beat clinical AI tools like OpenEvidence in all evaluations, matching Google Search AI Overview.

x.com/3d ago/3 min read
llmbenchmarkai
A laptop screen displays code with a sparse Mixture of Experts model diagram, symbolizing a Chinese lab's…
Open SourceBreakthrough
100

Chinese Lab's Free MoE Model Matches GPT-5.5 on Agentic Coding

A Chinese lab released an Apache-2.0 open-weights MoE model matching GPT-5.5 on agentic coding. This free model challenges proprietary AI's lead with sparse MoE architecture.

pub.towardsai.net/3d ago/3 min read/Widely Reported
open sourcecodingbenchmarks
Researchers in lab coats examining a glowing photonic chip under a microscope at Shanghai's new integrated photonic…
AI ResearchBreakthrough
95

China Launches Photonics Lab to Bypass US Chip Curbs on AI

China launched a photonics lab to bypass US chip curbs and develop energy-efficient AI computing using light instead of electrons.

scmp.com/3d ago/3 min read/Widely Reported
ai hardwarephotonicssemiconductors
OpenAI's Codex interface on a dark screen with code editor panels and a chat sidebar, likely showing autonomous…
Products & Launches
82

OpenAI Buys Ona to Give Codex Multi-Day Autonomous Coding

OpenAI acquired Ona (formerly Gitpod) to give Codex persistent cloud environments for autonomous coding tasks lasting hours or days, targeting Anthropic's Claude Code lead.

the-decoder.com/3d ago/3 min read
enterprise-aiai-agentsacquisitions
OpenAI researcher on social media announces Codex users get one free saved rate limit reset for Go, Plus, Pro, and…
Products & Launches
81

OpenAI Lets Codex Users Bank Rate Limit Resets

OpenAI lets Codex users save rate limit resets. One free saved reset for Go, Plus, Pro, Business tiers.

x.com/3d ago/3 min read
apicodexopenai
MiniMax M3 Exceeds Human Gold-Medal on Math Benchmarks via MaxProof
AI Research
99

MiniMax M3 Exceeds Human Gold-Medal on Math Benchmarks via MaxProof

MiniMax's M3 exceeded human gold-medal on math benchmarks via MaxProof, but no scores or details were disclosed.

x.com/3d ago/3 min read/Multi-Source
reasoningmathematicsai
Bar chart with two groups labeled 'Rejected' and 'Accepted' comparing prediction scores, with a ROC curve inset…
AI Research
72

Clinical LLM Rejection Predictor Hits AUROC 0.719 in 4.5-Month Study

Clinical LLM rejection predictor achieves AUROC 0.719 in 4.5-month study using deployment-specific context to forecast user rejection before response generation.

arxiv.org/3d ago/3 min read
llm evaluationarxivhealthcare
Jeff Bezos stands beside a glowing Prometheus logo on a dark tech stage, addressing investors at a funding event
Funding & BusinessBreakthrough
100

Bezos' Prometheus Closes $12B Round at $41B Valuation

Jeff Bezos' Prometheus raised $12B at $41B valuation, totaling $18.2B with no product. The compute-heavy startup targets physical-world AI but faces skepticism.

the-decoder.com/4d ago/3 min read/Widely Reported
ai startupsjeff bezosfunding
Two people in business attire shake hands in a modern office with server racks in the background, symbolizing a…
Startups
90

OpenAI Acquires Cloud Startup Ona to Power Agent Infrastructure

OpenAI acquired cloud startup Ona to support AI agent infrastructure, two days after a $6.6B raise. The deal targets enterprise reliability gaps as OpenAI pivots to B2B.

bloomberg.com/4d ago/3 min read/Widely Reported
enterpriseaibusiness

Lung-R1-14B Tops EMR Diagnosis with Knowledge Graph-Guide…

AI Research
92

Lung-R1-14B Tops EMR Diagnosis with Knowledge Graph-Guided RL

Lung-R1-14B scored 4.3583 on EMR diagnosis, beating 20 systems using a 59K-node knowledge graph and RL-constrained reasoning.

arxiv.org/4d ago/3 min read/Widely Reported
researchaihealthcare

SVoT Boosts MLLM Spatial Reasoning by 65% via RL-Verified…

AI Research
88

SVoT Boosts MLLM Spatial Reasoning by 65% via RL-Verified Visual Chains

SVoT uses RL to verify MLLM spatial reasoning states, achieving up to 65% accuracy gains on OOD tests across five domains including Pacman and Gather.

arxiv.org/4d ago/3 min read/Widely Reported
spatial-reasoningresearchreinforcement-learning
A sleek AI interface on a dark screen displays code analysis and exploit generation results, with a glowing network…
AI ResearchBreakthrough
84

Anthropic: Mythos Preview Builds Working Exploits in Hours, Not Weeks

Anthropic's Mythos Preview AI built 8 working exploits from Firefox and Windows kernel patches within hours. The first exploit was ready 18 days before the patched Firefox shipped.

the-decoder.com/5d ago/3 min read
claudeanthropicai security
Windows Hyper-V Manager interface showing a new virtual machine setup with Claude Desktop spawning a 1.8 GB VM on launch
Products & Launches
83

Claude Desktop Spawns 1.8 GB Hyper-V VM on Every Windows Launch

Claude Desktop spawns a 1.8 GB Hyper-V VM on every Windows launch due to 2,689 stale session files, consuming 11% of RAM.

github.com/5d ago/3 min read/Multi-Source
anthropicbugsclaude-desktop