Aravind Srinivas · lab leaders · perplexity ☆ FOLLOW 41 topic(s) over 14 day(s) — architecture, agents, inference. on X ↗
CARD NO. 2026-10-05 New models achieve top rankings across AI evaluation benchmarks evaluation · 15 voicesChatGPT adds native support for building and deploying MCP servers directly system_design · 12 voicesNvidia introduces the Open Agent Safety Platform for secure sandboxing agents · 9 voices
CARD NO. 2026-10-04 Google launches Gemini 4 Argon and Anthropic releases Claude Sonnet 5.5 evaluation · 14 voicesNvidia introduces Open Agent Safety Platform and ASR fine-tuning tutorials agents · 12 voicesChatGPT platform adds capability to host and deploy MCP servers system_design · 11 voicesFireworks AI introduces Ember-1 built on Kimi K3 with reduced token usage inference · 10 voicesGemini 4 Argon supports 1M output tokens while Claude Fable 5.5 rolls out inference · 9 voices
CARD NO. 2026-10-03 NVIDIA introduces Open Agent Safety Platform to secure AI agent execution agents · 12 voicesChatGPT Sites adds support for hosting and deploying MCP servers system_design · 9 voicesPerformance evaluations of Claude 5.5 and Fable 5.5 reveal reduced hallucination rates evaluation · 8 voices
CARD NO. 2026-10-02 NVIDIA launches Open Agent Safety Platform with industry partners system_design · 11 voicesDeepSeek releases desktop package and Pixel Canary model emerges agents · 9 voices
CARD NO. 2026-10-01 OpenAI releases GPT-6.1 and patches vision bugs across GPT-6 models architecture · 21 voicesNVIDIA Introduces Open Agent Safety Platform Combining OpenShell and Sentry agents · 9 voicesClaude Opus 5.5 improves accuracy and reduces hallucination rates evaluation · 7 voices
CARD NO. 2026-09-30 Grok 4.7 claims top ranking on the AA cyber index evaluation · 22 voicesNvidia introduces Open Agent Safety Platform with over 100 partners agents · 14 voices
CARD NO. 2026-09-29 Claude Opus 5.5 introduced with 40% lower operating costs architecture · 22 voicesNVIDIA introduces Open Agent Safety Platform with OpenShell and Sentry agents · 11 voicesOpenAI prepares for DevDay and expands capabilities for Astra agents · 10 voicesOpus 5.5 released with distillation and reinforcement learning from a teacher model training · 4 voices
CARD NO. 2026-09-28 Anthropic introduces Claude Opus 5.5 model architecture · 19 voicesOpenAI and xAI announce model lineups and benchmark improvements evaluation · 11 voicesIntroduction of the training and distillation approach for Opus 5.5 training · 4 voicesvLLM adds day-one support for the multimodal MiMo-V2.6 model family inference · 4 voices
CARD NO. 2026-09-27 Anthropic launches Claude Opus 5.5 with 40% lower operational costs architecture · 22 voicesOpenAI introduces GPT-6 Sol and GPT-6 Luna models architecture · 19 voicesNew research explores distillation and reinforcement learning for agents training · 4 voicesvLLM adds day-one support for Xiaomi MiMo-V2.6 and DiffusionGemma-Jev inference · 4 voices
CARD NO. 2026-09-26 OpenAI expands its ecosystem with GPT-6 Sol and Luna models architecture · 16 voicesAnthropic releases Claude Opus 5.5 amid tense White House negotiations architecture · 12 voicesTesseract creative suite launches with direct integration for Astra AI agents agents · 11 voicesvLLM and SGLang add day-zero support for new Xiaomi and GLM models inference · 3 voices
CARD NO. 2026-09-25 Anthropic releases the faster and cheaper Opus 5.5 model architecture · 11 voices
CARD NO. 2026-09-24 Astra launches Astra for Law alongside GPT-6 Sol and Luna models inference · 14 voicesAnthropic releases Claude Opus 5.5 featuring optimized token efficiency and performance architecture · 10 voicesOptimizing inference systems and accelerating execution for large language models inference · 3 voicesAlibaba announces large-scale AI infrastructure plans and RecreationWorld sandbox system_design · 3 voices
CARD NO. 2026-09-21 Nvidia expands AI hardware partnerships for Windows and enterprise system_design · 7 voices