https://www.seventnews.com/en/articles/opencode-hit-40m-arr-by-betting-developers-want-choices-not-lock-in SevenTnewS en 2026-07-24T21:54:54.011490+02:00 How an Anthropic ban backfired and drove OpenCode to $40M ARR https://www.seventnews.com/en/articles/claude-opus-5-matches-near-frontier-performance-at-half-the-cost-with-a-deliberate-blind-spot-on-cyber SevenTnewS en 2026-07-24T19:39:18.672675+02:00 Claude Opus 5 nearly matches Fable 5 across benchmarks, with cybersecurity as a deliberate blind spot https://www.seventnews.com/en/articles/microsofts-in-house-ai-models-cut-gpu-costs-up-to-89-in-production SevenTnewS en 2026-07-24T19:38:28.268015+02:00 The 89% GPU cost cut that changes the math for every AI company https://www.seventnews.com/en/articles/why-the-best-document-parser-now-fits-in-800-million-parameters SevenTnewS en 2026-07-24T19:19:36.643000+02:00 The 0.8B model that beat every pipeline document parser at its own game https://www.seventnews.com/en/articles/googles-cyber-fine-tune-finds-vulnerabilities-that-claude-opus-46-misses SevenTnewS en 2026-07-24T17:00:00+02:00 Google's cheap fine-tune found vulnerabilities that Claude Opus 4.6 missed https://www.seventnews.com/en/articles/hugging-face-pulls-the-infrastructure-lever-not-the-model-lever SevenTnewS en 2026-07-24T15:00:00+02:00 Hugging Face isn't fighting the model war. It's making deployment boring. https://www.seventnews.com/en/articles/graph-learning-models-arent-tested-on-data-this-ugly-a-new-benchmark-changes-that SevenTnewS en 2026-07-24T15:00:00+02:00 Your graph model breaks on dirty data. A new benchmark shows exactly where. https://www.seventnews.com/en/articles/a-new-framework-helps-ai-agents-remember-what-they-did-five-minutes-ago SevenTnewS en 2026-07-24T14:00:00+02:00 The one framework that lets AI agents remember what they did 47 steps ago https://www.seventnews.com/en/articles/langgraph-as-a-business-process-framework-not-an-ai-benchmark SevenTnewS en 2026-07-24T13:00:00+02:00 LangGraph is not an AI benchmark. It's a process framework that knows when to say no. https://www.seventnews.com/en/articles/what-deployment-reveals-about-agentic-systems-that-benchmarks-miss SevenTnewS en 2026-07-24T11:00:00+02:00 What breaks when AI agents leave the lab: a production reality check https://www.seventnews.com/en/articles/elevenlabs-just-got-a-serious-challenger-and-its-open-source SevenTnewS en 2026-07-24T11:00:00+02:00 A 68% win rate just made ElevenLabs look vulnerable https://www.seventnews.com/en/articles/four-personalities-one-answer-why-heterogeneous-reasoning-just-beat-the-best-ai-on-humanitys-hardest-test SevenTnewS en 2026-07-24T11:00:00+02:00 Four minds, one answer: why AI that thinks differently beat the biggest models at humanity's hardest test https://www.seventnews.com/en/articles/alibabas-happyoyster-lets-you-generate-explorable-3d-worlds-from-a-single-image SevenTnewS en 2026-07-24T10:39:00+02:00 Alibaba's HappyOyster builds a drivable 3D world from a single photo, and the catch is worth watching https://www.seventnews.com/en/articles/a-cnn-that-learns-to-like-and-dislike-things-just-like-you SevenTnewS en 2026-07-24T09:00:00+02:00 A neural network that learns to like and dislike things, just like you https://www.seventnews.com/en/articles/a-language-models-scrambled-neurons-just-recreated-a-strokes-effects-on-speech SevenTnewS en 2026-07-24T09:00:00+02:00 An AI model's scrambled neurons just recreated a stroke's exact damage to speech https://www.seventnews.com/en/articles/qwen-cloud-bundles-its-models-into-a-single-subscription-plan-cuts-team-pricing SevenTnewS en 2026-07-24T09:00:00+02:00 Qwen Cloud just made 40% of your AI costs vanish, and it works with Claude Code https://www.seventnews.com/en/articles/gemma-4-hits-300-million-downloads-making-open-weight-the-new-normal-for-ai-development SevenTnewS en 2026-07-24T08:45:40.029686+02:00 Gemma 4 hit 300 million downloads. The math that broke the cloud model is why. https://www.seventnews.com/en/articles/high-reconstruction-scores-dont-mean-your-ai-explanations-are-true SevenTnewS en 2026-07-24T08:00:00+02:00 Your AI thinks it explained itself. A new test proves it was lying the whole time. https://www.seventnews.com/en/articles/new-ling-30-flash-model-lands-on-opencode-for-free SevenTnewS en 2026-07-24T07:48:27.310531+02:00 OpenCode adds Ling 3.0 Flash for free, betting token efficiency wins the AI coding war https://www.seventnews.com/en/articles/sakana-ai-launches-fugu-ultra-11-a-router-that-picks-the-model-for-you SevenTnewS en 2026-07-24T07:38:08.040079+02:00 The router that hides which AI model runs your prompt, and why that matters https://www.seventnews.com/en/articles/the-missing-piece-in-ai-coding-assistants-why-memory-makes-the-difference-between-tool-and-partner SevenTnewS en 2026-07-24T00:00:00+02:00 The AI coding partner that finally remembers what you told it last week https://www.seventnews.com/en/articles/an-openai-model-breached-hugging-face-during-a-cyber-eval-and-neither-lab-noticed-first SevenTnewS en 2026-07-23T22:12:57.988274+02:00 OpenAI's own model escaped, breached Hugging Face, and nobody caught it first https://www.seventnews.com/en/articles/google-wants-gemini-to-run-your-side-hustle-the-problem-is-you-still-have-to-do-the-work SevenTnewS en 2026-07-23T20:35:00+02:00 Google's five-step side hustle guide skips the one step that actually makes money https://www.seventnews.com/en/articles/andrew-ngs-open-worker-lands-on-your-desktop-not-in-your-browser SevenTnewS en 2026-07-23T20:01:23.735757+02:00 OpenWorker is a framework that builds your work documents, not your chat history https://www.seventnews.com/en/articles/orienspace-takes-gravity-1-to-sea-as-chinese-commercial-launch-enters-batch-production SevenTnewS en 2026-07-23T19:00:00+02:00 China's first deep-sea rocket launch just made land sites obsolete for satellite internet https://www.seventnews.com/en/articles/why-alibabas-qwen25-omni-is-a-bigger-deal-than-its-7b-size-suggests SevenTnewS en 2026-07-23T17:00:00+02:00 The 7B model that beats bigger ones, runs on a phone, and changes the AI cost equation https://www.seventnews.com/en/articles/alibabas-happyoyster-generates-interactive-3d-worlds-from-a-single-prompt SevenTnewS en 2026-07-23T15:37:02.380708+02:00 Alibaba's HappyOyster generates explorable 3D worlds from a single sentence https://www.seventnews.com/en/articles/how-ai-tracking-data-is-reshaping-the-way-football-clubs-recruit-and-scout SevenTnewS en 2026-07-23T15:32:45.781983+02:00 The off-ball revolution: how AI tracking data is rewriting scouting https://www.seventnews.com/en/articles/the-high-school-race-where-the-only-fuel-is-sunlight SevenTnewS en 2026-07-23T15:00:00+02:00 630 miles, zero gas, and a car that breaks in your own parking lot https://www.seventnews.com/en/articles/lg-quietly-pushes-mcafee-ads-through-its-monitors-via-windows-update SevenTnewS en 2026-07-23T15:00:00+02:00 LG is using Windows Update to turn its monitors into ad machines https://www.seventnews.com/en/articles/alibabas-qwen-image-30-turns-text-rendering-into-a-productivity-argument SevenTnewS en 2026-07-23T10:38:00+02:00 Alibaba's latest image model doesn't want to paint. It wants to print your newspaper. https://www.seventnews.com/en/articles/why-vision-language-models-need-a-visual-relay-window-to-stop-hallucinating SevenTnewS en 2026-07-23T09:00:00+02:00 The three-stage rhythm that stops AI from seeing things that aren't there https://www.seventnews.com/en/articles/alibabas-qwen-coder-qoder-beats-cursor-on-its-own-test-and-slashes-token-use-by-145 SevenTnewS en 2026-07-23T07:44:00+02:00 Alibaba built a coding model that learns from its own users. The numbers are hard to ignore. https://www.seventnews.com/en/articles/what-a-game-engine-can-teach-ai-about-keeping-its-world-straight SevenTnewS en 2026-07-23T07:09:00+02:00 The game engine rule that generative models keep forgetting https://www.seventnews.com/en/articles/sakana-ai-brings-nvidia-nemotron-into-its-fugu-multi-agent-orchestration-system SevenTnewS en 2026-07-23T02:24:00+02:00 Nemotron meets Fugu: Sakana AI's bet that open models win as a swarm, not alone https://www.seventnews.com/en/articles/when-every-ai-agent-speaks-a-different-skill-language-nacos-builds-the-rosetta-stone SevenTnewS en 2026-07-23T00:00:00+02:00 Your AI coding agents are running old playbooks. Nacos fixes that. https://www.seventnews.com/en/articles/the-livebench-leaderboard-is-a-study-in-diminishing-returns SevenTnewS en 2026-07-22T22:51:53.375563+02:00 The LiveBench top four are separated by 2.2 points. The cost difference is brutal.