SpaceXAI has launched Grok 4.5, a new AI model developed in partnership with Cursor, designed to handle complex, long-running tasks across various domains including software engineering, legal, and financial services. The model boasts significant improvements in token efficiency, reportedly using about 4.2 times fewer output tokens than Anthropic's Opus 4.8 on the SWE Bench Pro. Grok 4.5 also achieved the top rank on Harvey's Legal Agent Benchmark and is priced at $2/M input and $6/M output tokens, with availability through SpaceXAI's console and Cursor. AI
IMPACT Sets new SOTA on legal benchmarks and offers significant token efficiency gains, potentially lowering costs for complex agentic tasks.
RANK_REASON Frontier-lab model release with system card. [lever_c_demoted from frontier_release: ic=2 ai=1.0]
- Anthropic
- ChatGPT
- Cursor
- GPT-Live-1
- GPT-Live-1 mini
- Grok 4.5
- OpenAI
- SpaceX AI
- Harvey's Legal Agent Benchmark
- NVIDIA GB300
- Opus 4.8
- SpaceXAI
- SWE Bench Pro
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →