https://tpsreport.news/news/agent-teams-vs-solo-agents-vibe-code-bench-vals-ai TPS — Tokens Per Second en 2026-10-11T15:50:46.963Z Vals AI: Agent teams cost 1.8x–5.1x more than solo agents, with one significant gain in four tests agent teams, multi-agent, Vals AI, Vibe Code Bench, GPT-6 Sol, Claude Opus 5.5, token cost, benchmark https://tpsreport.news/news/odyssey-3-world-model-public-preview TPS — Tokens Per Second en 2026-10-11T08:05:53.131Z Odyssey-3 world model enters free public preview; 14B base model, 66.1 Physics-IQ claim carries caveats Odyssey, Odyssey-3, world model, generative video, robotics, Physics-IQ, WorldMark, research preview https://tpsreport.news/news/microsoft-decision-1-qwen3-5-9b-decision-model TPS — Tokens Per Second en 2026-10-10T15:05:49.633Z Microsoft releases Decision-1, a Qwen3.5-9B-based model for classification and routing, at $0.042 per 1M input tokens Microsoft, Decision-1, decision model, Qwen3.5-9B, small language models, OpenRouter, Microsoft Foundry, agent routing https://tpsreport.news/news/gemini-app-free-users-auto-model-flash-lite TPS — Tokens Per Second en 2026-10-10T01:35:44.917Z Google removes model picker for free Gemini app users, locking them to 'Auto' routing that defaults to Flash-Lite Gemini, Google, Gemini app, Flash-Lite, model routing, free tier, thinking levels, changelog https://tpsreport.news/news/alphafold-protein-folding-kohli-candido-latent-space TPS — Tokens Per Second en 2026-10-10T00:50:48.406Z DeepMind's Pushmeet Kohli and Biohub's Sal Candido Explain Why AlphaFold Didn't Solve Protein Folding AlphaFold, Google DeepMind, Biohub, protein folding, AI for science, Latent Space, Pushmeet Kohli https://tpsreport.news/news/anthropic-turns-off-live-internet-access-internal-evals TPS — Tokens Per Second en 2026-10-10T00:35:46.574Z Anthropic cuts internal evals off from the live internet after its AI agents exploited government sites Anthropic, AI safety, AI agents, reward hacking, alignment, evaluations, containment, OpenAI