xAI has released Grok 4.6, the latest iteration of its frontier large language model, scoring 61 on the Artificial Analysis Intelligence Index to match OpenAI's GPT-5.6 Sol. The model launched today and is available immediately in Cursor and Grok Build, with API access and partner integrations through OpenRouter, Vercel, and Cloudflare.

For the latest breaking AI news and analysis, visit AI Buzz Wire.

Frontier Intelligence at a Fraction of the Cost

According to Artificial Analysis, an independent benchmarking platform, Grok 4.6 gains five points over its predecessor Grok 4.5 on the Intelligence Index, bringing xAI back to the frontier tier alongside OpenAI. The model sits just behind Anthropic's Claude Opus 5 (63) and Claude Fable 5 (62), and ahead of Kimi K3.

What makes Grok 4.6 particularly notable is its pricing. Headline pricing remains unchanged from Grok 4.5 at $2 per million input tokens and $6 per million output tokens. That puts it more than 60% below Claude Opus 5 ($5/$25) and GPT-5.6 Sol ($5/$30). The measured cost per task is $0.84, placing it on the Intelligence versus Cost Pareto frontier.

Holding pricing flat across a generation is unusual at the frontier, where intelligence gains have typically been accompanied by price increases. Grok 4.6 delivers a five-point Intelligence Index gain at unchanged pricing.

Built for Long-Running Agents

Grok 4.6 builds on Grok 4.5 with what xAI describes as a particular focus on long-running agents and more ambitious interactive and visual work. The company says the model stays with complex tasks across many steps, whether researching a topic, analyzing information, working across a codebase, or turning an idea into a polished application.

The agentic benchmarks tell a compelling story. On GDPval-AA v2, a leading measure of real-world agentic knowledge work, Grok 4.6 achieves an Elo of 1753, behind only Claude Opus 5 and statistically on par with Claude Fable 5 and Qwen3.8 Max. It scores 50.7% on the tau-cubed Banking benchmark for multi-turn customer service, placing it in the top two alongside Qwen3.8 Max.

On CursorBench v3.2, Grok 4.6 scores 69.9%, ahead of both GPT-5.6 Sol (67.2%) and Grok 4.5 (66.7%). It posts 65.9% on DeepSWE v1.1 and 61.3% on FrontierCode v1.1 Extended.

Remarkable Turn Efficiency

One of the standout findings from Artificial Analysis is Grok 4.6's efficiency profile on AA-Briefcase, a private benchmark of long-horizon agentic knowledge work tasks. The model debuts with an Elo of 1577, placing it at the Fable 5 tier.

The efficiency numbers are striking. Grok 4.6 resolves tasks in approximately 53 turns and 0.5 billion input tokens on average, compared to roughly 103 turns and 2.0 billion input tokens for Claude Opus 5 max. Since long-horizon agentic work accumulates context rapidly, a model that reaches a comparable answer in half the turns and a quarter of the input tokens has a cost advantage well beyond its per-token pricing.

Training and Safety

Grok 4.6 underwent a longer supplemental training run than Grok 4.5, with curated model-generated data for reasoning and advanced technical concepts, high-quality engineering data, and an improved optimizer and training recipe. xAI used Grok 4.5 to regenerate SFT trajectories across reasoning efforts, agent harnesses, and domains including STEM, software engineering, and knowledge work.

The model carries a 500,000-token context window, unchanged from Grok 4.5. xAI reports that Grok 4.6 underwent its widest-ever suite of pre-deployment testing for capabilities and safeguard calibration, along with extensive post-deployment and third-party testing.

Availability and Pricing

Grok 4.6 is available today in Cursor and Grok Build. xAI is offering double included usage inside both platforms for the first week. Pricing starts at $2 per million input tokens and $6 per million output tokens, with a fast variant available at twice the price. Cache hits are discounted to $0.50 per million tokens.

The release continues a rapid cadence for xAI, with Grok 4.6 arriving just over a month after Grok 4.5. The model positions xAI as a serious contender in the frontier race, particularly for developers prioritizing agentic performance and cost efficiency.

Stay Ahead of AI

For continuous coverage of the AI industry's most important developments, bookmark AI Buzz Wire and never miss a breaking story.

Read more AI news