SpaceXAI Launches Grok 4.7: Low Prices, Heavy Token Use
SpaceXAI’s Grok 4.7 improves coding benchmarks and keeps API prices low, but heavy token consumption could raise the cost of completing real-world tasks. The post SpaceXAI Launches Grok 4.7: Low Prices, Heavy Token Use appeared first on TechRepublic .
SpaceXAI has launched Grok 4.7, its latest AI coding model designed to deliver strong performance without increasing base prices. However, the model's heavy token consumption could lead to higher costs for completing real-world tasks. Grok 4.7 was released on September 21 and builds upon its predecessor, Grok 4.6, using a larger base model and undergoing extensive reinforcement learning focused on complex, multi-step tasks.
Independent testing reveals that Grok 4.7 consumes significantly more output tokens than its predecessor and competitors, potentially leading to higher costs despite its competitive pricing. Starting at $2 per million input tokens and $6 per million output tokens, Grok 4.7's pricing remains unchanged from Grok 4.6. However, the model's higher token usage could offset these savings if the model needs more tokens to complete tasks.
Grok 4.7 demonstrates improved performance in coding benchmarks, scoring higher on CursorBench 4.0, Terminal-Bench 4.0, DeepSWE v1.1, and EEBench compared to Grok 4.6. It also shows enhanced conversational abilities and knowledge work capabilities due to its native understanding of the Grok Bot harness. Nonetheless, the model still lags behind selected competitors in certain professional knowledge-work and clinical benchmarks.
SpaceXAI has implemented new safeguards to enhance Grok 4.7's resistance to jailbreak attempts while limiting unnecessary refusals. The model scored 62.4% on LatchBio's biosafety benchmark and allowed 3.3% of risky dual-use prompts tested with HackerBench v0.3. Selected cybersecurity partners can access Grok 4.7 for red-team testing.
Developers will find Grok 4.7 particularly useful for complex tasks involving repositories, terminals, repeated tool calls, and extended coding sessions. Professionals can benefit from its focus on documents, presentations, and extended knowledge-work tasks. Grok 4.7 is available through Cursor, Grok Build, the Grok API, third-party coding tools, model routers, and cloud platforms.
However, its unpredictable benchmark results and high token consumption necessitate thorough testing and consideration of cost per successful task before implementation.
Written by urgent.news from TechRepublic's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
