Grok 4.5

On July 28, 2026, SpaceXAI announced that Grok 4.5 is now available in GitHub Copilot, expanding distribution to millions of developers using VSCode and GitHub products. Developers can select "Grok 4.5" from the model picker in cloud agents, the Copilot CLI, and the VSCode IDE; enterprises may need to enable the model This partnership announcement positions Grok 4.5 as a coding-focused model within GitHub's ecosystem, linking the launch coverage and benchmark results to a concrete developer-tooling integration. For more on coding performance details, SpaceXAI points readers back to the original Grok 4.5 announcement page. The integr

On July 22, 2026, SpaceXAI announced that Grok 4.5 now powers Grok across consumer surfaces: grok.com, X, and the Grok apps for iOS and Android. The post highlights Grok 4.5 as the company's most capable model yet, emphasizing better adherence to user questions, longer conversation context tracking, clearer answers, an SpaceXAI also detailed Grok 4.5's capabilities for end-to-end knowledge work, including building spreadsheets, turning outlines into slides and diagrams, drafting prose, and summarizing long PDFs. Integration with Microsoft 365 add-ins brings Grok directly into Word, Excel, PowerPoint, and Outlook. The post directs rea

SpaceXAI (xAI) officially announced Grok 4.5 on July 16, 2026, describing it as the company's smartest model built for coding, agentic tasks, and knowledge work. The model was trained alongside Cursor and across tens of thousands of NVIDIA GB300 GPUs, with heavy investment in data filtering, deduplication, quality scor SpaceXAI published detailed coding and engineering benchmarks for Grok 4.5. On DeepSWE 1.0 (Datacurve eval, run by AA), Grok 4.5 scored 62.0%, compared to Fable at 66.1% and GPT 5.5 at 64.31%. On Terminal Bench 2.1, Grok 4.5 reached 83.3%, slightly below Fable (84.3%) and GPT 5.5 (83.4%) but ahead of Opus 4.8 (78.9%).

FullStack Labs' August 14, 2026 article confirms SpaceXAI announced Grok 4.5 on July 8, 2026 as its most capable model so far, focused on coding, agents, and knowledge work, built on a new large-scale foundation model incorporating data from real developer activity. Elon Musk's launch posts on X described Grok 4.5 as r The piece emphasizes Grok 4.5's design for tools and systems handling code and long documents, with long-context reasoning, tool use, and stable behavior over real repositories and internal knowledge sources. It can read and modify large repositories, coordinate multi-step plans through tools or APIs, and process long

Tech Insider Ireland's July 27, 2026 coverage confirms xAI launched Grok 4.5 on July 8, 2026 into the coding market, with Elon Musk describing it on X as an Opus-class model that is faster, more token-efficient, and lower cost. The model shipped priced at $2 per million input tokens and $6 per million output tokens, un Access widened through summer 2026: EU API users on July 17, 2026, before reaching consumer surfaces (Grok.com, X, iOS, Android) on July 22, 2026 as SpaceXAI's main reasoning model. The launch framing positions Grok 4.5 as a direct shot at territory occupied by Anthropic's Claude Code, GitHub Copilot, and Cursor, with

LLM Stats' Grok 4.5 page presents composite scoring, benchmark rankings, and a blended price point of $2.19 per million tokens, placing the model at a composite LLM Stats Score of 45.4 on the aggregator's leaderboard. Performance-by-conversation-depth figures show scores ranging from 13.2 (turns 31+) to 15.4 (turns 11- The aggregator tracks Grok 4.5's performance stability through September 17, 2026, with capability-tier placement and cost-efficiency positioning against models like Gemma 4 E4B, GPT OSS 120B, DeepSeek-V4-Flash-0731, DeepSeek-V4.1-Flash, and GPT-6 Astra. Per-domain sigma movements show positive deltas in Chat (+1.25σ,

DataCamp's July 14, 2026 coverage confirms SpaceXAI released Grok 4.5 on July 8, 2026, as a proprietary mixture-of-experts reasoning model targeted at coding, agent tasks, and general office work, with Cursor co-training on trillions of tokens from developer and agent interactions. Grok 4.5 exposes a 500,000-token cont DataCamp groups SpaceXAI's launch claims into coding with agents, token use, office work, and tool support, noting that SpaceXAI's published charts show mixed benchmark results rather than a clean sweep of wins. The piece stresses that per-token rates tell only part of the cost story once reasoning, retries, and tool c

Medium/MLWorks' July 9, 2026 post confirms SpaceXAI launched Grok 4.5 on July 8, 2026, built on the V9 foundation model architecture, which is a ground-up redesign of xAI's model foundation rather than a fine-tune of the previous v8-small architecture. V9 runs at 1.5 trillion parameters, a 3x scale increase from v8-sma The article traces the Grok family trajectory: Grok 4 (July 2025, V8, 256K context), Grok 4 Fast (V8, 2M context), Grok 4.1 (coding-focused agentic release), Grok 4.20 (multi-agent 4-agent council, native 2M context), Grok 4.3 (500B parameters, V8, available via $300/month heavy tier), and Grok 4.5 (1.5T V9, with Curso