SpaceXAI launched Grok 4.5 on July 16, 2026, its most capable model for coding, agentic tasks, and knowledge work. The model was trained alongside Cursor and delivers competitive benchmark results while running at 80 tokens per second - faster than many flash models. Its token efficiency cuts output costs, with pricing set at $2 per million input tokens and $6 per million output tokens.
Benchmark performance
Grok 4.5 scores near the top on several coding benchmarks. On DeepSWE 1.0, it reached 62.0%, behind Fable (max) at 66.1% and GPT 5.5 (xhigh) at 64.31%. For SWE Marathon, it led with a 29.0% resolution rate, ahead of Opus 4.8 (max) at 26.0%. On Terminal Bench 2.1, it scored 83.3%, just behind Fable's 84.3% and GPT 5.5's 83.4%. The model resolves SWE Bench Pro tasks with an average of 15,954 output tokens - about 4.2 times fewer than Opus 4.8 (max) at 67,020 tokens.
Training and efficiency
SpaceXAI's latest Grok (from xAi) model was trained across tens of thousands of NVIDIA GB300 GPUs with heavy investment in data filtering, deduplication, and quality scoring. The team scaled reinforcement learning on hundreds of thousands of tasks, focusing on multi-step software engineering and per-token intelligence. The result is a model that handles real engineering tasks with fewer steps and lower cost per task. Grok 4.5 serves at 80 tokens per second, and its token efficiency means it often finishes coding tasks in under half the output tokens of comparable leading models.
Coding and office capabilities
For coding tasks, Grok 4.5 can build functional applications from a single prompt, such as a solar system simulation using three.js with realistic orbits and a modern HUD. In Grok Build, it is the default model for complex Excel models that pull web research, use multi-sheet formulas, and leave notes for future reference. The model also creates PowerPoint diagrams with native shapes and writes clear prose in Word.
Pricing and availability
Grok 4.5 costs $2 per million input tokens and $6 per million output tokens, undercutting many competitors on price while offering roughly twice the token efficiency. It is available now in Grok Build, in Cursor on all plans, and via the SpaceXAI API. Developers can start with a free tier in Grok Build and Cursor for a limited time.
Why this matters for IT and development professionals
Grok 4.5 brings down the cost and time to resolve complex coding and agentic tasks. Its token efficiency means fewer API calls and lower bills for the same work. For product development teams, the ability to generate working apps from minimal prompts can accelerate prototyping. IT teams can use Grok Build's Excel and Office capabilities to automate reporting and documentation with less manual effort.
Your membership also unlocks:
