Musk "copied" Zhipu
- Key Point: xAI released Grok 4.5, claiming it achieves comparable intelligence to the Anthropic Opus 4.7 at an output price roughly 80% lower, adopting a high cost-performance strategy. This approach aligns with strategies of Chinese open-source AI companies, putting pressure on the market for high-end closed-source models.
- Key Factors:
- xAI Pricing Strategy: Grok 4.5's output token pricing is set at $6 per million tokens, less than a quarter of Anthropic Opus 4.7 ($25), while claiming comparable capabilities.
- Token Efficiency Advantage: xAI states Grok 4.5's token efficiency is twice that of other leading models, significantly reducing enterprise inference costs—constituting its core selling point.
- Intensified Industry Competition: Closed-source models like Anthropic's that rely on high pricing face pressure on both capability and cost dimensions. OpenAI is also set to release GPT 5.6 this week.
- Parallel to Chinese Strategy: Analysts comment this move resembles Chinese AI companies' approach of capturing market share through price competition. Previously, Zhipu's GLM-5.2 approximated Anthropic's flagship model with 82% lower pricing.
- Coding Agent Strategy: The strategic value of xAI's acquisition of Cursor lies in gaining access to massive real-world programming data, which positions it to challenge Claude Code's leading position in the agent market.
Original Author: Bao Yilong
Original Source: Wall Street CN
xAI releases Grok 4.5, challenging Anthropic with "Opus-level" capabilities, its cost-performance strategy directly mirrors the Chinese open-source model approach.
On Wednesday, Elon Musk's xAI released its latest model, Grok 4.5, claiming to achieve intelligence on par with Anthropic's flagship model at an output token price less than one-fifth of Anthropic's.
This pricing strategy closely aligns with Zhipu AI's recent GLM-5.2, which matched closed-source frontier models at an 82% lower price point, leading market analysts to observe that Silicon Valley's top labs are beginning to adopt the cost-performance playbook of Chinese open-source vendors. Musk himself characterized the release on X:
Grok 4.5 is an Opus-level model, but faster, more token-efficient, and cheaper.
He further stated that internal evaluations show Grok 4.5's overall capabilities are roughly on par with Opus 4.7, but "significantly faster." The combination of capability, speed, and low cost forms the core competitive logic of this model.
According to xAI's published pricing, Grok 4.5 costs $2 per million input tokens and $6 per million output tokens. In comparison, Anthropic's Opus 4.7 charges $5 and $25 per million input and output tokens respectively, a stark price difference.
If this pricing strategy delivers on its claimed capabilities, it will directly impact the procurement decisions of enterprise AI customers. Meanwhile, OpenAI is scheduled to release its latest flagship model, GPT 5.6, this Thursday, making this week a dense window for AI model releases.
Cost-Performance Strategy, Token Efficiency as Core Selling Point
xAI disclosed in a blog post that Grok 4.5's token efficiency is twice that of other leading models.
If this efficiency is realized in practical applications, it will directly reduce inference costs for enterprises, a highly attractive proposition for business users increasingly concerned about token expenditures in AI procurement.
For comparison, in OpenAI's tiered pricing system, the highest-tier 'Sol' version charges $5 and $30 per million input and output tokens, respectively, while the lowest-tier 'Luna' version charges $1 and $6.
Grok 4.5's pricing is significantly lower than its main competitors, and within Musk's narrative framework, it corresponds to capabilities nearing Anthropic's flagship model.
Analyst "Serenity" commented on X, stating that xAI's move somewhat resembles what Chinese AI vendors have been doing: competing through price competition and user adoption.
Just a few weeks ago, Zhipu AI's open-source model GLM-5.2, priced approximately 72% to 82% lower, scored 74.4 on the FrontierSWE programming benchmark, less than one percentage point behind Anthropic Opus 4.8's 75.1 and surpassing GPT-5.5's 72.6.
In a subsequent report, JPMorgan summarized this phenomenon, stating that mature intelligence continues to compress pricing, but frontier capability upgrades can still support premiums, leading to structural divergence in the monetization logic of the AI model layer.
The arrival of Grok 4.5 further complicates this competitive landscape: as top US labs begin to enter the market with pricing strategies similar to open-source models, institutions like Anthropic, which rely on high-priced closed-source models, will face pressure in both capability and cost dimensions.
Programming Agent Market, Cursor Acquisition Logic Emerges
In terms of product positioning, xAI describes Grok 4.5 as a "workhorse model" for daily knowledge tasks, covering areas such as programming and application development, office documents, research, and content writing. This echoes the strategic logic of its recent acquisition of the code editing tool, Cursor.
Jukan, an analyst at investment research firm Citrini Research, noted in a comment that at the recent ICML conference, he heard a notable bullish thesis for Grok: Claude Code currently leads the programming agent market, not solely due to superior model quality, but because of the massive user base gained from its first-mover advantage. More users generate more real-world coding data, which in turn feeds back into model quality, creating a flywheel effect.
Jukan believes the core value of xAI's Cursor acquisition lies precisely here:
Cursor likely has a larger real-world user base and code dataset than OpenAI Codex. If xAI can effectively train on and utilize this data, surpassing Codex might just be a matter of time.
