Grok 4 Fast Ranks #2 on OpenRouter, Nears 100B Tokens a Day with Grok Code Fast 1 Leading

News Batra
23 Sep, 2025

 Grok 4 Fast Hits #2 on OpenRouter; Joins Grok Code Fast 1 at the Top

Nears 100B Tokens a Day


September 23, 2025 — xAI’s new model Grok 4 Fast surged to the #2 spot on the OpenRouter leaderboard, processing nearly 100 billion tokens a day. It joins its sibling model Grok Code Fast 1 in the top slot, marking a notable moment in the ongoing LLM (large language model) arms race.

Key Metrics & Why They’re Impressive

  • Token usage: On a single day, Grok 4 Fast handled about 97.8B tokens via OpenRouter’s ranked usage stats. That figure puts it clearly in the #2 position behind Grok Code Fast 1.

  • Leaderboard standings:

    1. Grok Code Fast 1 — #1
    2. Grok 4 Fast — #2
    3. Next challenger (Claude Sonnet 4 from Anthropic) trails behind in terms of token volume.

  • Feature highlights of Grok 4 Fast:

    • 2 million token context window — allows much larger inputs (e.g. long documents, extended conversation) to be processed more coherently.
    • Pricing & efficiency: It is significantly cheaper per token than its predecessor Grok 4 under many usage conditions. xAI claims about 40% fewer “thinking tokens” on average for comparable tasks, which equates to a steep reduction in cost to deliver similarly strong benchmark performance.

What Grok Code Fast 1 Brings & Why It’s Still #1

  • Specialization in coding tasks: Grok Code Fast 1 is tailored for programming workflows — agentic coding, fast completions, likely fewer delays.
  • Massive adoption early on: Within days of its release, it captured a lion’s share of usage from developer tools (e.g. Kilo Code) and other platforms.
  • Speed + cost advantage: Reports put its throughput speed high and pricing favorable, making it attractive for both hobbyists and enterprise users wanting coding assistance without high fees.

Why This Matters: Implications for AI, Users & Competitors

  • Developer choice & competition: Models that are fast, cheap, and strong enough are lowering barriers for developers. Grok 4 Fast’s large context window plus efficiency make it more usable in varied real-world scenarios. Grok Code Fast 1’s coding-focus means serious competition for models like Claude Code or GPT’s code offerings.

  • Cost-efficiency pressures: With users seeing models that deliver similar benchmark performance at lower costs, other providers will face pressure to optimize token usage, context windows, and pricing.

  • Benchmark vs. real usage: Leaderboard dominance is one thing; integrating stability, safety, and generalization across real tasks is another. While usage numbers are impressive, evaluating quality (error rates, hallucinations, tool integration) will be crucial.

  • Model architecture trends: The shift toward large context windows, unified reasoning / non-reasoning modes, and token-efficiency reflects broader trends in LLM development. Grok 4 Fast is riding these trends well.


Grok 4 Fast Ranks #2 on OpenRouter, Nears 100B Tokens a Day

Grok Code Fast 1 and Grok 4 Fast Dominate the Leaderboard Together

September 23, 2025 — The AI race just heated up as Grok 4 Fast secured the #2 spot on the OpenRouter leaderboard, processing close to 100 billion tokens in a single day. This rapid rise puts it shoulder to shoulder with Grok Code Fast 1, which continues to lead at the #1 position. Together, these models from xAI are setting a new standard for speed, efficiency, and large-scale adoption.

Why Grok 4 Fast Matters

Unlike traditional models, Grok 4 Fast is designed to handle massive context windows of up to 2 million tokens, making it capable of analyzing long conversations, full documents, and complex workflows in one go. What’s more impressive is its token efficiency — reports suggest it uses 40% fewer “thinking tokens” compared to its predecessor, reducing costs while maintaining benchmark strength.

This makes Grok 4 Fast particularly attractive for businesses and developers who need scalable AI power without breaking budgets.

Grok Code Fast 1: Still the Leader

While Grok 4 Fast is turning heads, Grok Code Fast 1 continues to dominate as the #1 ranked model on OpenRouter. Optimized for programming tasks, it delivers lightning-fast completions and reliable coding support. Its cost-effective pricing and adoption across developer platforms like Kilo Code have fueled its rapid growth.

For developers, this means better access to an AI model that is not only accurate but also practical for day-to-day coding assistance.

Implications for the AI Industry

The dominance of Grok Code Fast 1 and Grok 4 Fast highlights a few big trends:

  • Efficiency wins: Users are flocking to models that balance strong performance with affordable usage.
  • Bigger context, bigger potential: Large context windows open doors for real-world applications like research, law, data analysis, and multi-document reasoning.
  • Competitive pressure: Rivals such as Claude, GPT, and Gemini will need to push harder to match both the cost savings and scale of adoption that Grok models are showing.

Final Takeaway

With Grok Code Fast 1 holding the top spot and Grok 4 Fast rising to #2, xAI has cemented its presence as a leader in the LLM ecosystem. Nearly 100 billion tokens processed daily is no small feat, and it shows just how quickly users are adopting these models.


Here are the best verified news and source links for your article 👇




Next Post Previous Post
No Comment
Add Comment
comment url