--- title: Grok 4.5 launches as xAI's flagship coding and agentic model description: xAI launched Grok 4.5 on July 16, positioning it as its top model for coding, agentic tasks, and knowledge work. It topped the Long-Horizon Terminal-Bench... canonical: https://daily.dev/posts/grok-4-5-launches-as-xai-s-flagship-coding-and-agentic-model-rksbgbgmq twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Grok 4.5 launches as xAI's flagship coding and agentic model | daily.dev og:description: xAI launched Grok 4.5 on July 16, positioning it as its top model for coding, agentic tasks, and knowledge work. It topped the Long-Horizon Terminal-Bench... og:url: https://daily.dev/posts/grok-4-5-launches-as-xai-s-flagship-coding-and-agentic-model-rksbgbgmq og:image: https://api.daily.dev/og/posts/RKsBGbgMQ.png og:image:alt: Grok 4.5 launches as xAI's flagship coding and agentic model og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Grok 4.5 launches as xAI's flagship coding and agentic model **[Collections](https://daily.dev/sources/collections)** · 2 min read · 3 upvotes · 0 comments ## Summary xAI launched Grok 4.5 on July 16, positioning it as its top model for coding, agentic tasks, and knowledge work. It topped the Long-Horizon Terminal-Bench leaderboard across 46 tasks with a mean reward of 0.505, beating Claude Sonnet 5, Opus 4.8, and GPT-5.6 Sol. The benchmark tests long multi-step workflows averaging 231 agent steps and 85 minutes per task. On cybersecurity benchmarks, Grok 4.5 offers the best price-performance ratio — 10x cheaper than Sol while matching Kimi K3 performance. Grok Build, xAI's coding CLI now powered by Grok 4.5, supports plan mode, parallel subagents, MCP plugins, and a fullscreen terminal UI, and is available to try for free. ## Content xAI launched Grok 4.5 on July 16 as its strongest model yet for coding, agentic tasks, and knowledge work. It's rolling out across platforms now, though high demand has caused some speed issues for early users. ## Benchmark results The headline number comes from Long-Horizon Terminal-Bench, where Grok 4.5 ranked first across 46 tasks covering software engineering, research reproduction, scientific computing, and multimodal analysis. Every model ran through the same Terminus-2 harness with a single 90-minute attempt per task. The average run took around 231 agent steps, 85 minutes, and 9.9 million tokens — so this isn't a "write a function" benchmark. It tests whether a model can hold context, debug, recover from errors, and actually finish a long workflow. Grok 4.5's results: - Mean reward: 0.505 (highest in the field) - Tasks solved at ≥0.95 threshold: 13/46 - Perfect scores (reward = 1.0): 7/46 It finished ahead of Claude Sonnet 5, Opus 4.8, Fable 5, and GPT-5.6 Sol. On SWE-bench it also shows strong results, and separate cybersecurity benchmarks put it at the best price-performance ratio in that category — 10x cheaper than Sol, 5.7x cheaper than Opus 5, and 2.2x cheaper than Kimi K3, while matching Kimi-level performance. Sol still leads on raw capability. One caveat worth noting: the Terminal-Bench tests ran Grok 4.5 inside the shared Terminus-2 harness, not inside Grok Build itself. So the results say something about the underlying model, not necessarily that Grok Build beats Claude Code or Codex in practice. ## Grok Build Grok Build is xAI's coding CLI, now running on Grok 4.5. It can build complete functional apps from simple prompts and is positioned for professional engineering work. Current features include: - Plan mode - Parallel subagents with separate context windows - Mouse-interactive fullscreen terminal UI - Skills, plugins, and MCP support You can try it free: ```bash curl -fsSL https://grok.build/install.sh | bash ``` Image generation and creative tasks are also supported, but the focus is clearly on code quality and long-running agentic workflows. --- Tags: [#llm](https://daily.dev/tags/llm), [#ai-assisted-development](https://daily.dev/tags/ai-assisted-development), [#agentic-ai](https://daily.dev/tags/agentic-ai), [#grok](https://daily.dev/tags/grok) [View this post on daily.dev](https://daily.dev/posts/grok-4-5-launches-as-xai-s-flagship-coding-and-agentic-model-rksbgbgmq) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Grok 4.5 launches as xAI's flagship coding and agentic model","url":"https://daily.dev/posts/grok-4-5-launches-as-xai-s-flagship-coding-and-agentic-model-rksbgbgmq","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/grok-4-5-launches-as-xai-s-flagship-coding-and-agentic-model-rksbgbgmq"},"datePublished":"2026-07-27T21:21:59.829Z","dateModified":"2026-07-27T21:22:38.853Z","description":"xAI launched Grok 4.5 on July 16, positioning it as its top model for coding, agentic tasks, and knowledge work. It topped the Long-Horizon Terminal-Bench...","isAccessibleForFree":true,"articleSection":"Collections","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Collections","logo":"https://media.daily.dev/image/upload/s--fk_6ycEi--/f_auto,q_auto/v1780996001/logos/collections?_a=BAMAMiWQ0","url":"https://daily.dev/sources/collections"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/grok-4-5-launches-as-xai-s-flagship-coding-and-agentic-model-rksbgbgmq","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":3},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,ai-assisted-development,agentic-ai,grok","timeRequired":"PT2M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Collections","item":"https://daily.dev/sources/collections"},{"@type":"ListItem","position":3,"name":"Grok 4.5 launches as xAI's flagship coding and agentic model"}]} ```