https://tpsreport.news/news/timestral-209m-time-series-forecasting TPS — Tokens Per Second en 2026-10-19T17:20:57.665Z Mistral releases Timestral-209M, a time series model claiming top zero-shot scores on three benchmarks Mistral AI, Timestral, time series, forecasting, foundation model, open weights, GIFT-Eval, fev-bench https://tpsreport.news/news/nvidia-nemotron-3-5-super-vl-120b-a12b-nvfp4 TPS — Tokens Per Second en 2026-10-19T15:36:02.637Z NVIDIA releases Nemotron 3.5 Super VL NVFP4: 120B-parameter, 12B-active vision-language model with 1M context NVIDIA, Nemotron, vision-language, NVFP4, mixture-of-experts, Mamba-2, open-weights, quantization https://tpsreport.news/news/gemini-4-argon-agents-beat-minecraft-autonomously TPS — Tokens Per Second en 2026-10-18T12:50:47.798Z Four Gemini 4 Argon agents beat Minecraft autonomously in 27 hours, killing the Ender Dragon at $1,620 cost Gemini 4 Argon, Google DeepMind, Minecraft, AI agents, multi-agent systems, autonomous agents, research https://tpsreport.news/news/taste-bench-agent-long-term-decisions-gpt-5-6-sol-59-7 TPS — Tokens Per Second en 2026-10-18T09:05:49.318Z Taste-Bench: Best AI model scores 59.7% on long-term agent decisions, where random guessing yields 25% Taste-Bench, benchmark, AI agents, Microsoft, GPT-5.6 Sol, SWE-bench Pro, Qwen3.6-27B, long-horizon agents https://tpsreport.news/news/tencent-hy-memory-30b-a3b-memory-writer TPS — Tokens Per Second en 2026-10-18T04:35:56.971Z Tencent releases Hy-Memory-30B-A3B, a 3B-active model that writes agent memory for Mem0 and MemOS Tencent, Hunyuan, Hy-Memory, long-term memory, Mem0, MemOS, mixture-of-experts, open weights https://tpsreport.news/news/openai-python-sdk-3-35-0-ultrafast-agents-service-tier TPS — Tokens Per Second en 2026-10-17T23:20:44.390Z OpenAI Python SDK 3.35.0 adds ultrafast service tier for Agents and typed Live sideband events openai, openai-python, sdk, changelog, agents, realtime, service-tier, websockets