xAI Launches Grok 4.5, Undercutting Opus 4.8 and GPT-5.5 on Price
xAI released Grok 4.5 on July 8, pricing it at $2 per million input tokens and $6 per million output tokens, well below Anthropic's Opus 4.8 and OpenAI's GPT-5.5, while posting benchmark scores close to both on coding tasks.
xAI made its new flagship model, Grok 4.5, publicly available on July 8, calling it an 'Opus-class' model that trades a small benchmark gap for a much lower price. The model runs on V9, xAI's ninth-generation, 1.5 trillion parameter architecture, and its training pipeline was supplemented with data from Cursor to improve coding performance.
- Priced at $2 per million input tokens and $6 per million output tokens, versus Opus 4.8's $5/$25 and GPT-5.5's roughly $5/$30
- Scores 62.0% on DeepSWE 1.0, ahead of Opus 4.8's 55.75%
- Scores 83.3% on Terminal Bench 2.1, just behind GPT-5.5's 83.4% and Claude Fable 5's 84.3%
- Trails on DeepSWE 1.1 at 53%, versus GPT-5.5's 67% and Fable 5's 70%
- Available now in the Grok app, in Cursor on all plans, and via the xAI console
Faster, more token-efficient and lower cost than Opus, and roughly comparable on quality. - Elon Musk
The mixed benchmark picture, ahead on some coding suites, behind on others, suggests xAI is betting that a roughly 60-70% price cut against Opus 4.8 and GPT-5.5 will matter more to buyers than a clean sweep of leaderboards. Compare it yourself in the coding arena.