xAI Unveils Grok 4.5 for Programming and Agents

🤖 xAI has introduced Grok 4.5, optimized for programming and agentic tasks.

The model delivers 80 tokens/sec and consumes 4.2x fewer tokens on SWE Bench Pro tasks than Claude 3 Opus. Grok 4.5 competes in performance with GPT-5.5 and Opus 4.8.

🌍 The emergence of Grok 4.5 sets a new standard for context efficiency, which is critical for scaling autonomous AI engineers.

👤 You can use Grok 4.5 via Cursor or API for cheaper and faster coding. The model is particularly strong in Rust and C/C++.

Source 1: https://x.ai/news/grok-4-5 Source 2: https://docs.x.ai/