이미지: X — 뉴스 앰프 화면 갈무리
Summary
- Grok 4.6 pricing was revealed at $2 per million input tokens and $6 for output
- That's below Claude Opus 5 ($5/$25) and GPT-5.6 Sol ($5/$30)
- DeepSeek 4 Pro GA reportedly set an even lower price than that
- Grok 4.6 가격
- 입력 $2 / 출력 $6 (100만 토큰당)
- Grok 4.6 Fast 가격
- 입력 $4 / 출력 $12
- Claude Sonnet 5 가격
- 입력 $2 / 출력 $10
- Claude Opus 5 가격
- 입력 $5 / 출력 $25
- GPT-5.6 Sol 가격
- 입력 $5 / 출력 $30
What a single price sheet reveals
xAI's Grok 4.6 has come in at $2 per million input tokens and $6 for output. Compared with Claude Opus 5 ($5 input/$25 output) and GPT-5.6 Sol ($5 input/$30 output) on the same table, that's anywhere from half to a fifth of the price. Claude Sonnet 5 matches Grok 4.6's $2 input price but charges more for output at $10. Chubby, an account that tracks AI news, shared the table and noted that "price is becoming more important than anything, even performance." Chubby added that DeepSeek 4 Pro GA undercut even this price.
Performance already top-tier — now it's a price war
Grok 4.6 was already ahead in the performance debate. In the Artificial Analysis Intelligence Index results released on August 12, Grok 4.6 scored 61 points, placing it in the same top tier as GPT-5.6 Sol and up 5 points from its predecessor Grok 4.5. On GDPval-AA v2, a practical-task benchmark, it recorded an Elo of 1753, ranking second behind Claude Opus 5 (1849), and it's reportedly already been integrated and tested in the coding tool Cursor. The key point of this pricing table is that the price came down without sacrificing performance.
Price competition isn't new in the AI model market. Since running large language models costs money per token processed, it makes sense to pick the cheaper model if it delivers the same answer. The catch has been the assumption that "cheaper models perform worse" — and Grok 4.6 is challenging that assumption by matching GPT-5.6-level intelligence scores while cutting price. On top of that, China's DeepSeek reportedly priced its latest model, DeepSeek 4 Pro GA, even lower than Grok 4.6, pushing the floor of the price war down another notch.
So what changes now
Where model selection once came down to a single benchmark score, it's now shifting to a "cost per point" calculation. For developers and companies running high-volume API calls, a difference of a few dollars per token can reshape an entire monthly bill. With Grok 4.6 delivering GPT-5.6/Claude-level performance at less than half the price, and DeepSeek pricing even lower still, the focus of model selection is moving away from "which model is smartest" and toward "which model can do the same job the cheapest."
