subagent-review.grok.me

Official notes on subagents: Claude Opus 5, GPT-5.6, Kimi K3, and Muse Spark 1.2 — including system-card crew evals for score, latency, and tokens.

mio.grok.me

MIO wraps AI models with an explicit layer for observation, grounding, uncertainty, coherence, evidence, and action.

island.grok.me

Argus · 望台 Argus 望台 进入控制台 Model Availability Monitor 用真实请求 核验每个模型。 为 NewAPI / OneAPI / OpenAI Compatible / Anthropic / Gemini 中转站建立统一探针。 导入 all-api-hub 与 CC Switch 配置，定时巡检，记录延迟、健康度、费用与格式合格率。 开始监测 登录后导入站点 真实探测 走 /v1/chat/completions 等协议，而不是只看目录是否列出。 五分钟切片 按站点与模型记录可用性、延迟、质量分与调用费用。…

gold-stone-eagle-eagle.grok.me

Yiying Hu，Jarsy 联合创始人兼 COO。MIT 校友，曾任职 Uber、Square、Ripple。

frontier-models.grok.me

Interactive 3D cost–score–speed dominance staircase for frontier language models.

bold-fern-cliff-zippy.grok.me

Public instrument tracking the evolution of the Colossus supercomputer complex. Official, observed, and inferred — clearly tagged.

apex-ever-terra-zenith.grok.me

Discover which AI model fits the job — chat, reasoning, coding, research, website updates, image, video, voice, and speed.

tae01.grok.me

Token Arbitrage Engine — route inference across a pool of LLM keys by idle capacity, reset urgency, cost, and trust.

claude.grok.me

Step-by-step Claude starter course: prompts, iteration, templates, and safe daily use.