Made with Grok Imaginehttps://t.co/ik8qteEEtI ...

Long press on any image to turn it into a video!Then customize the prompt to create whatever you can Imagine.My prompt here was “add a boyfriend and they transition into muppets 😂 https://t.co/nfNjT8ZK3mEmily (@IamEmily2050):Grok Imagine https://t.co/nTDD86gpLl ...

RT notsure (@kelenthir)@beffjezos Upgrade is a euphemism https://t.co/uWpejhUSmV ...

RT Dogan Ural (@doganuraldesign)Show me your cool Grok Imagine shots!Try the new text-to-image model, it's upgradedPure Grok image + video ↓ https://t.co/V0UIjBjDFI ...

RT Eric Jiang (@veggie_eric)hold on tight https://t.co/8Worw9nfPz ...

Update your @Grok app for the latest capabilitieshttps://t.co/u2y4RZSsODDogeDesigner (@cb_doge):Another upgrade for Grok Imagine. It’s now at the point where it’s almost impossible to tell what’s real and what’s AI.Update your Grok app! https://t.co/ZIqtDZJkiN ...

RT James Lucas (@JamesLucasIT)@elonmusk Aeneas fleeing TroyMade with Grok Imagine https://t.co/UpVDDwPKIa ...

Haha exactly 😂Yun-Ta Tsai (@YunTaTsai1):Designing an inference chip for robots is actually very difficult.In data centers each chip is bathed in jacuzzi and babysat by nannies. If they died it would be hot swapped by one of their clones.The fault rate of GPUs in datacenter is actually quite high. Industrial average ...

Model Capabilities - Grok 4 Fast 拥有 200 万 token 的超大上下文窗口 [1] - 这一窗口允许加载大量文档或代码库，作为单个提示进行查询和分析 [1] - Grok 4 Fast 速度极快 [1] Industry Impact - 行业将受益于更大的上下文窗口，从而可以处理更复杂的任务 [1] - 行业可以更高效地分析大量数据 [1]

Model Update - xAI released a significant update for Grok-4-Fast [1] - Grok4 is considered the best model currently [1] - Grok-4-Fast should be used with Grok4-Heavy and Grok-Code-Fast-1 for code-related tasks [1] Performance Improvement - Reasoning Mode completion rate increased from 77.5% to 94.1% [1] - Non-Reasoning Mode completion rate increased from 77.9% to 97.9% [1] Features - 2 million context window, unified modes, and state-of-the-art (SOTA) performance [1] - Multimodal capabilities (text + images) with integrated web/X search [1] API & Pricing - API pricing: $0.20 per million input tokens, $0.50 per million output tokens [1] - SuperGrok/Premium+ offers unlimited access [1]