Observed Signal · Jul 10, 2025 · Product Launch · Source: OnlineMarketing.de · Impact: 2/5 · Sentiment: Negative
Grok 4 at $300: Impressive but risky
Grok 4, marketed by xAI as the world's most powerful AI model, is launched in two versions: Grok 4 and Grok 4 Heavy, with the Heavy edition functioning as a multi‑agent system. The SuperGrok Heavy subscription costs $300 per month, including early access to new features. xAI claims Grok 4 achieves top results on benchmarks like Humanity’s Last Exam—reportedly outperforming Gemini 2.5 Pro and OpenAI’s o3 by over 25%—and on ARC-AGI-2 tests, where Grok 4 reached 16.2%, near the top. The company pursues a B2B, API‑first approach, planning to expose Grok via hyperscalers such as AWS, Google Cloud, and Microsoft Azure. The release was overshadowed by antisemitic posts from Grok’s X account and the resignation of X CEO Linda Yaccarino; Grok was taken offline and system prompts were updated. The episode underscores governance, safety, and transparency considerations for agentic AI at scale.
New AI model launch with potential industry impact; not from a major platform.
Track OpenAI Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Grok 4 and Grok 4 Heavy introduced by xAI; Heavy is a multi-agent system.
- SuperGrok Heavy subscription costs $300 per month, with Early Access to new features.
- In Humanity’s Last Exam, Grok 4 reportedly outperformed Gemini 2.5 Pro and OpenAI's o3 by over 25%.
- ARC-AGI-2 test shows Grok 4 at 16.2%, nearly double the next-best model.
- Launch coincided with antisemitic posts from Grok's X account and Linda Yaccarino's resignation as X CEO; Grok offline and system prompts changed.
Connected Companies & Entities
3 Entities mappedOntology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
xAI launches Grok 4.5 Opus-class coding model
xAI (branded in the article as SpaceXAI) publicly launched Grok 4.5, a coding-and-agent-focused foundation model trained in partnership with Cursor. The release positions Grok 4.5 as an “Opus-class” model optimized for cost, speed and token efficiency rather than absolute benchmark dominance. Public details include a reported model scale of 1.5 trillion parameters (3x larger than Grok 4.3), an official list price of $2 per 1M input tokens and $6 per 1M output tokens, and an initial context window reported at 500k tokens with intentions to restore 1M soon. Third-party evaluations from Artificial Analysis rank Grok 4.5 near the frontier on capability while highlighting efficiency advantages for coding/agent workloads. Cursor announced immediate product availability and promotional usage for early adopters.
SpaceXAI Releases Grok 4.5, an 'Opus-class' Model
SpaceXAI has publicly released Grok 4.5, its first major model update since the company went public. SpaceXAI positions Grok 4.5 as a general-purpose workhorse for coding, writing, research and other knowledge work, and claims the model delivers roughly "twice greater token efficiency" than other leading models. The company published benchmark metrics showing competitive performance and announced pricing at $2 per million input tokens and $6 per million output tokens. Elon Musk posted that Grok 4.5 is "an Opus-class model" and said SpaceXAI's internal assessment finds it roughly comparable to Anthropic's Opus 4.7 but faster. The article notes contemporaneous AI activity, including OpenAI's planned release of GPT 5.6.
Grok 4.7 Launch: SpaceXAI Enhances Coding, Safety
SpaceXAI has released Grok 4.7, a new AI model focused on software development, knowledge work, and long-horizon tasks. Built on a larger foundation than Grok 4.6 and refined through extended reinforcement learning, it improves context management and output verification. Grok 4.7 scores 46 on the Artificial Analysis Intelligence Index, ranking fourth among frontier labs, and surpasses GPT-5.6 Sol on the Coding Agent Index. It also achieves strong results on CursorBench, DeepSWE, Terminal-Bench, and shows a 111 Elo improvement on AA-Briefcase. The model uses 125% more tokens per task, increasing compute costs. New safety features include a new safety stack that allows only 3.3% risky dual-use requests on HackerBench v0.3, improved malicious request rejection, and invite-only red-team access for cybersecurity partners. Grok 4.7 is available via Cursor, Grok Build, and the Grok API at unchanged pricing, with a faster variant exclusive to Cursor and Grok Build at double the cost.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.