xAI Launches Grok 4.6 With Frontier Performance at Lower Price
Elon Musk’s xAI has been pushing its Grok model family as a lower-cost alternative to frontier systems from OpenAI and Anthropic. Grok 4.6 builds on its predecessor, with much of the performance gain attributed to post-training using supervised fine-tuning, or SFT, and reinforcement learning. The model targets coding, research and long-running agent workflows, areas increasingly important as companies move beyond chatbots toward software that can execute complex, multi-step tasks.
xAI released Grok 4.6 on Aug. 12, 2026. The model scored 61 on the Artificial Analysis Intelligence Index, five points above Grok 4.5 and level with OpenAI’s GPT-5.6 Sol. Its DeepSWE 1.1 score rose to 65.9% from 54%, while CursorBench 3.2 reached 69.9%. Standard API access costs $2 per million input tokens and $6 per million output tokens, compared with GPT-5.6 Sol’s $5 and $30, respectively. A higher-speed Grok variant is priced at twice the standard rate.
All Coverage
1 original reportsThe Backstory
The history behind this eventxAI Launches Grok Bot, Rolls Out Grok 4.6
xAI is extending Grok beyond conversational AI into autonomous task execution with Grok Bot, an agent product that gives each bot its own cloud computer. The setup is designed to let agents work across apps, continue running around the clock and function more like digital employees than chat assistants. The launch matters because major AI developers are increasingly competing to control workflows and software actions, not merely answer prompts, raising the stakes in enterprise automation and coding tools.
xAI opened Grok Bot for public testing, touting the ability to operate 24 hours a day, while Elon Musk said Grok 4.6 would arrive later in the week. Subsequent reports said xAI formally released Grok 4.6, with benchmark results described as matching GPT-5.6 and stronger coding capabilities made available through Cursor. The rapid shift from preview to release gives developers an early route to test the model in production-style coding workflows.
xAI Launches Grok 4.5 to Challenge Anthropic
Elon Musk-founded AI startup xAI has been working to compete with industry heavyweights including OpenAI and Anthropic. As large language models gain traction among businesses and developers, computational efficiency and deployment costs have become key competitive battlegrounds. xAI’s new model aims to disrupt the market with greater token efficiency and lower pricing while directly challenging rival flagship products.
xAI launched the Grok 4.5 large language model and made its API available on July 16, 2026. Built on a 1.5-trillion-parameter base model and trained using GB300 systems, it offers a 500K context window and twice the token efficiency of comparable models. Grok 4.5 is priced at $2 per million input tokens and $6 per million output tokens, positioning it against Anthropic’s Opus 4.7.
Subscribe to Mark Radar Weekly
Every Friday, the week's strongest signals in your inbox. Unsubscribe anytime.