Elon Musk's xAI and Google DeepMind have both unveiled new artificial intelligence models designed to compete on price and long-running task capabilities, signaling a major shift in how US AI firms are battling for enterprise customers. Grok 4.6, released by xAI, and Gemini 3.7 Flash, Google's latest offering, represent a strategic pivot away from pure performance benchmarks toward affordability and practical agent work.

What Are Grok 4.6 and Gemini 3.7 Flash Designed to Do?

Grok 4.6 builds on xAI's previous model with a particular focus on long-running agents and more ambitious interactive and visual work. The model is engineered to handle complex tasks across many steps, whether researching a topic, analyzing information, working across a codebase, or turning an idea into a polished application. xAI highlights that Grok 4.6 is especially strong at turning a broad product idea into a working first version, researching unfamiliar domains, structuring applications, implementing core interactions, and refining results through multiple rounds of feedback.

Google's Gemini 3.7 Flash is billed as the company's most intelligent workhorse model yet for coding and agents. The model delivers substantial improvements across software engineering, knowledge work, and web development workflows. For knowledge-dense fields like finance, law, and biosciences, Gemini 3.7 Flash delivers improved reasoning and accuracy, according to Google.

How Are These Models Competing on Price?

Cost has become a critical battleground for US AI firms. Both xAI and Google are aggressively undercutting their rivals to retain cost-conscious customers who are switching to cheaper alternatives from Chinese competitors. Grok 4.6 costs $0.84 per task, which is exactly the same price as Chinese firm Moonshot's Kimi K3. Gemini 3.7 Flash carries an introductory price of half the original 3.6 Flash cost per million tokens, positioning it as a lower-cost option for enterprises.

"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," said Ivan Zhou, AI Research Manager at Databricks.

Ivan Zhou, AI Research Manager at Databricks

Even leading US labs such as OpenAI and Anthropic are releasing cheaper models to fight for cost-conscious customers, according to reporting by the Financial Times cited in the source material. This pricing pressure reflects a broader market dynamic where Chinese AI firms are undercutting Western competitors on cost, forcing established players to choose between margin and market share.

Where Do These Models Rank Against Competitors?

According to benchmark data from Artificial Analysis, Grok 4.6 ranks ninth on a competitive leaderboard of top AI models, while Gemini 3.7 Flash ranks seventh. The leaderboard shows that Anthropic's Claude Fable 5 Max Effort and OpenAI's GPT-5.6 Sol Max Effort still lead across most benchmarks, but the gap is narrowing as xAI and Google improve their offerings.

• Top-Tier US Models: Claude Fable 5 Max Effort (Anthropic) and GPT-5.6 Sol Max Effort (OpenAI) continue to dominate the leaderboard rankings
• Chinese Competitors: Moonshot's Kimi K3 and Alibaba's Qwen 3.8 Max are gaining ground, particularly on cost metrics that appeal to enterprise buyers
• Emerging Contenders: Grok 4.6 and Gemini 3.7 Flash represent the next wave of competition, combining improved capabilities with aggressive pricing strategies

Elon Musk claimed on the platform X that "Grok 4.6 is objectively #1 when considering intelligence, speed and cost," though this claim is arguable rather than objective. The competitive landscape shows that while Grok 4.6 and Gemini 3.7 Flash may not lead on pure intelligence benchmarks, they are competitive on the combined metric of capability, speed, and affordability that enterprise customers increasingly prioritize.

How Are Google and xAI Restructuring to Compete?

Both companies are making organizational changes to accelerate their AI efforts. Google is currently overhauling its AI division, Google DeepMind. Demis Hassabis stepped down from his role as CEO of Google DeepMind to become Chair and Chief Scientist of Google's parent company Alphabet. In a message announcing Demis's departure, Sundar Pichai, CEO of Google and Alphabet, wrote: "We have to accelerate all this work and stay focused on the AI frontier".

Sundar Pichai, CEO of Google and Alphabet

According to Reuters reporting cited in the source material, Google Co-founder Sergey Brin has urged key AI staff to go all in on the company's Gemini model, likely in a bid to compete with the top offerings from OpenAI and Anthropic. Meanwhile, xAI, led by CEO Elon Musk, launched Grok Bot, which it describes as "AI teammates you can give real work to," signaling a shift toward practical, agent-based applications rather than pure conversational AI.

Steps to Understand the Enterprise AI Market Shift

• Benchmark Performance: Compare models not just on raw intelligence scores but on the combination of capability, speed, and cost per task, which better reflects what enterprises actually care about
• Pricing Transparency: Look at per-token or per-task pricing to understand how US firms are responding to Chinese competition and what that means for your AI infrastructure costs
• Agent Capabilities: Evaluate whether new models can handle multi-step tasks and long-running workflows, as this is becoming a key differentiator between consumer-grade and enterprise-grade AI
• Safety and Guardrails: Review how companies are calibrating safety features alongside capability improvements, as this affects real-world deployment risk

xAI has highlighted that Grok 4.6's safeguards have been improved and calibrated in line with the model's capabilities, addressing previous concerns about the model's safety features. This reflects a broader industry recognition that as AI models become more capable, their safety mechanisms must evolve in tandem.

The release of Grok 4.6 and Gemini 3.7 Flash marks a turning point in the AI market. Rather than competing solely on raw intelligence or feature richness, xAI and Google are now fighting for enterprise adoption by offering high-quality models at prices that undercut both each other and Chinese rivals. This price-driven competition suggests that the AI market is maturing from a frontier-research phase into a more commoditized enterprise software phase, where cost efficiency and practical task completion matter as much as benchmark scores.