Grok 4.5
Grok 4.5 is xAI/SpaceXAI's July 2026 model for coding, agentic work, and knowledge work. The official launch says it is available in Grok Build, Cursor, and the xAI API with the `grok-4.5` model ID.
Grok 4.5 matters because it is both a new frontier-model candidate and a harness-specific launch: xAI positions it for coding and agents, Cursor says it was trained jointly with SpaceXAI, and the public pricing makes it practical to compare against Claude, GPT, and Composer routes by task cost.
The xAI launch page provides the model ID, price, API example, availability, EU caveat, and coding/office-agent positioning. Cursor confirms product availability and the Cursor training relationship. X/Twitter and Artificial Analysis posts are useful demand and benchmark-discussion signals, but model facts should come from xAI, Cursor, and current provider docs.
- Run difficult coding-agent tasks in Cursor, Grok Build, or the xAI API.
- Compare cost-per-task against Claude, GPT, Composer, and GLM routes.
- Use effort levels or harness routing to match task difficulty.
- Evaluate agentic office workflows that combine coding, spreadsheets, documents, and slides.
xAI describes Grok 4.5 as a model for coding, agentic tasks, and office-style work. The launch page lists the `grok-4.5` API model ID, says it is available in Grok Build, Cursor, and the xAI console, and publishes API pricing of $2 per million input tokens and $6 per million output tokens.
- Availability: Grok Build, Cursor, and the xAI API console at launch.
- Pricing: $2 per million input tokens and $6 per million output tokens in the xAI API launch page.
- Regional caveat: xAI says EU availability is not present at launch and is expected later in July 2026.
Cursor frames Grok 4.5 as a model trained together with SpaceXAI and released across Cursor desktop, web, iOS, CLI, and SDK surfaces. That makes the model a practical coding-agent candidate, not just a generic chat model. Cursor also states Composer remains available for smaller, faster everyday coding work.
Treat xAI and Cursor as the factual sources for availability, model ID, and product positioning. Use Artificial Analysis, X posts, Hacker News, or Reddit only to discover comparison questions such as cost-per-task, coding benchmark fit, Cursor usage limits, EU availability, and whether Grok 4.5 should replace or complement Composer, Claude, or GPT routes.
The runtime layer where Grok 4.5 is evaluated for coding, tools, effort levels, and task cost.
The API-selection concept for comparing model IDs, pricing, context, and provider surfaces.
Trace review is the right way to compare Grok 4.5 against other agent models on real tasks.
Grok 4.5 FAQ
Page-level questions for Grok 4.5.
What is the Grok 4.5 model ID?+
The xAI launch page shows `grok-4.5` in its API example. Verify current access and regional availability in the xAI console or the provider surface you plan to use before adding it to production routing.
Is Grok 4.5 available in Cursor?+
Yes. Cursor announced Grok 4.5 availability across Cursor desktop, web, iOS, CLI, and SDK surfaces. Cursor also says Composer remains available for smaller everyday coding tasks, so teams should compare both routes by task size, cost, latency, and result quality.
How much does Grok 4.5 cost on the xAI API?+
The xAI launch page lists Grok 4.5 at $2 per million input tokens and $6 per million output tokens. Treat Cursor pricing, free launch usage, and any fast variants as separate product-surface details to verify from Cursor or the active provider.