https://tpsreport.news/news/gemini-4-google-teases-larger-base-model TPS — Tokens Per Second en 2026-07-26T20:20:45.210Z Google Confirms Gemini 4 Training Underway, Says Next-Gen Model Needs a Larger Base Model to Compete at the Frontier Gemini 4, Google, Google DeepMind, Sundar Pichai, Gemini 3.5 Pro, frontier models, TPU https://tpsreport.news/news/cursor-agent-swarm-planner-worker-architecture TPS — Tokens Per Second en 2026-07-26T15:21:05.750Z Cursor's New Agent Swarm Architecture Splits Planning and Coding to Cut Costs by Up to 15x Cursor, Anysphere, agent swarm, AI coding, Composer 2.5, Kimi K2.5, multi-agent systems https://tpsreport.news/news/microsoft-mage-flow-4b-image-editing TPS — Tokens Per Second en 2026-07-26T13:50:48.584Z Microsoft Releases Mage-Flow: Compact 4B Image Generation and Editing Models Matching Systems 5-8x Larger microsoft, image-generation, image-editing, diffusion-model, open-source, text-to-image, huggingface https://tpsreport.news/news/altman-white-house-openai-model-hugging-face-hack TPS — Tokens Per Second en 2026-07-26T12:50:44.815Z Altman to Brief White House on Unreleased OpenAI Model That Autonomously Hacked Hugging Face OpenAI, Sam Altman, AI policy, White House, AI safety, agentic AI, Hugging Face https://tpsreport.news/news/claude-opus-5-arc-agi-3-record TPS — Tokens Per Second en 2026-07-26T09:50:49.348Z Claude Opus 5 Scores 30.2% on ARC-AGI-3, Nearly 4x the Previous Record Claude Opus 5, Anthropic, ARC-AGI-3, benchmark, AI reasoning, GPT-5.6 Sol, ARC Prize https://tpsreport.news/news/microsoft-fara1-5-27b-web-agent TPS — Tokens Per Second en 2026-07-25T11:36:09.750Z Microsoft Releases Fara1.5-27B, a 27B Vision-Only Web Browsing Agent with 262K Context Microsoft, Fara1.5, computer use agent, Qwen3.5, open weights, web automation, multimodal https://tpsreport.news/news/claude-opus-5-prompt-injection-zero-percent TPS — Tokens Per Second en 2026-07-25T10:50:51.665Z Anthropic's Claude Opus 5 Hits 0% Prompt Injection Success Rate in Browser Agent Tests, With Defenses Enabled Anthropic, Claude Opus 5, prompt injection, AI agents, AI security, Gray Swan https://tpsreport.news/news/claude-opus-5-benchmarks-beat-fable-5-cost TPS — Tokens Per Second en 2026-07-25T09:50:57.117Z Claude Opus 5 Scores 61 on Intelligence Index, Beats Fable 5 on Cost Across Most Benchmarks Claude Opus 5, Anthropic, Artificial Analysis, Epoch AI, benchmarks, LLM pricing, Claude Code, reasoning models https://tpsreport.news/news/claude-opus-5-fable-level-performance-half-price TPS — Tokens Per Second en 2026-07-25T07:36:03.922Z Anthropic Ships Claude Opus 5, Claims Near-Fable Performance at Half the Price Anthropic, Claude Opus 5, model release, benchmarks, Epoch AI, Artificial Analysis, agentic AI