xAI and Google have released competing AI models designed to handle complex, multi-step tasks at lower costs, signaling an intensifying battle for enterprise customers as both firms chase rivals Anthropic and OpenAI. Grok 4.6, developed by Elon Musk's xAI, emphasizes long-running agents and visual reasoning, while Google's Gemini 3.7 Flash focuses on coding and knowledge-intensive work at half the price of its predecessor.

What Makes These New Models Different from Earlier Versions?

Grok 4.6 builds directly on its predecessor, Grok 4.5, with a particular emphasis on sustaining work across many steps. xAI tested the model on projects designed to "stretch its range" and "ability to sustain work over many steps," according to the company. The model excels at turning broad product ideas into working prototypes, researching unfamiliar domains, structuring applications, implementing core interactions, and refining results through multiple rounds of feedback.

Gemini 3.7 Flash, meanwhile, represents Google's latest effort to compete in the enterprise market. The model delivers substantial improvements across software engineering, knowledge work, and web development workflows. Google emphasizes that the new version provides a noticeably improved developer experience compared to its predecessor, Gemini 3.6 Flash, better adapting to roadblocks, clarifying intent when needed, and following instructions with greater fidelity.

How Are Pricing and Cost Becoming the Competitive Battleground?

Cost has emerged as a critical differentiator in the AI market. Gemini 3.7 Flash is priced at half the original cost per million tokens of its predecessor, making it an attractive option for cost-conscious enterprises. Grok 4.6 costs $0.84 per task, which matches the pricing of Chinese competitor Moonshot's Kimi K3 model. This aggressive pricing reflects broader industry pressure as leading US labs like OpenAI and Anthropic release cheaper models to retain customers switching to cut-price alternatives from Chinese rivals.

"Until now, getting high-quality answers from complex enterprise data has been expensive at scale. Models like Gemini 3.7 Flash are changing that by delivering better intelligence at dramatically lower cost," stated Ivan Zhou, AI Research Manager at Databricks.

Ivan Zhou, AI Research Manager at Databricks

For knowledge-dense fields like finance, law, and biosciences, Google argues that Gemini 3.7 Flash delivers improved reasoning and accuracy, making it particularly valuable for enterprises handling sensitive, complex information.

Where Do These Models Rank Among Global AI Competitors?

Despite the new releases, both models face stiff competition from established leaders. According to recent benchmarking data, the competitive landscape includes:

• Top Performers: Claude Fable 5 Max Effort (Anthropic) and GPT-5.6 Sol Max Effort (OpenAI) continue to lead across most benchmarks
• Chinese Competitors: Moonshot's Kimi K3 and Alibaba's Qwen 3.8 Max rank ahead of Grok 4.6 in current evaluations
• Grok 4.6 Position: xAI's model ranks ninth on recent benchmarks, below Google's Gemini 3.7 Flash, which ranks seventh

Elon Musk claimed on X (formerly Twitter) that "Grok 4.6 is objectively #1 when considering intelligence, speed and cost," though this assertion is disputed by independent benchmarking data.

How Are xAI and Google Restructuring to Compete?

Both companies are making organizational changes to sharpen their competitive edge. Google is overhauling its AI division, Google DeepMind, with Demis Hassabis stepping down as CEO to become Chair and Chief Scientist of parent company Alphabet. Google CEO Sundar Pichai stated the company must "accelerate all this work and stay focused on the AI frontier". Additionally, Google Co-founder Sergey Brin has urged key AI staff to prioritize the company's Gemini model to compete with top offerings from OpenAI and Anthropic.

Sundar Pichai

xAI, meanwhile, is expanding Grok's capabilities beyond the base model. The company launched Grok Bot, described as "AI teammates you can give real work to," positioning the technology for enterprise adoption. This move signals xAI's ambition to move beyond model development into practical enterprise applications.

What Are the Key Differences in Agent Capabilities?

Agent capabilities have become central to both releases. An AI agent is a system that can autonomously perform tasks, make decisions, and take actions over extended periods with minimal human intervention. Grok 4.6 is particularly strong at handling complex tasks that require sustained reasoning across many steps. The model can research topics, analyze information, work across codebases, and turn ideas into polished applications or work products.

Gemini 3.7 Flash is billed as Google's "most intelligent workhorse model yet for coding and agents," suggesting it is optimized for developers building agent-based systems and for enterprises deploying AI to handle routine coding and knowledge work.

Steps to Evaluate These Models for Enterprise Use

• Assess Your Cost Requirements: Compare pricing per task or per token based on your expected usage volume; Grok 4.6 and Gemini 3.7 Flash both offer competitive rates at $0.84 per task and half-price per token respectively
• Test Agent Capabilities: Run pilot projects on your specific use cases, such as code generation, data analysis, or multi-step research tasks, to evaluate how well each model sustains performance across extended workflows
• Evaluate Domain-Specific Performance: For knowledge-intensive fields like finance, law, or biosciences, benchmark Gemini 3.7 Flash's reasoning and accuracy against your current solutions
• Consider Integration Needs: Examine how each model integrates with your existing enterprise systems and whether Grok Bot or Google's agent frameworks align with your infrastructure

What Safety and Legal Concerns Surround Grok?

xAI has faced scrutiny over Grok's safety guardrails. The company noted that Grok 4.6's safeguards have been improved and calibrated in line with the model's capabilities. However, BBC News reported in March that xAI is facing a lawsuit from teenagers who claim the company facilitated child pornography by allowing the creation of sexually explicit images of them. Additionally, a federal judge recently ruled that xAI cannot move a privacy lawsuit filed by a Grok user to Texas, keeping the case in its original jurisdiction.

These legal challenges underscore the regulatory and safety pressures facing AI companies as they scale their models and user bases. xAI's emphasis on improved safeguards in Grok 4.6 appears to be a direct response to these concerns.

What Does This Competition Mean for Enterprise Customers?

The release of Grok 4.6 and Gemini 3.7 Flash reflects a broader shift in the AI market toward cost-competitive, capable models that can handle real enterprise work. As xAI and Google compete with Anthropic and OpenAI for market share, enterprise customers benefit from lower prices, improved agent capabilities, and more specialized models tailored to specific industries and workflows. The aggressive pricing and focus on multi-step reasoning suggest that AI vendors are moving beyond raw benchmark performance to practical, deployable systems that can reduce operational costs while improving productivity.