Grok 4.5 surpasses all leading models in invoice processing
xAI
xAI's Grok 4.5 model has outperformed all major competitors in invoice processing benchmarks. The model achieved state-of-the-art results on multiple accounting tasks, including data extraction and table interpretation.
According to xAI, Grok 4.5 set a new state-of-the-art on the Invoice Processing Benchmark, surpassing GPT-4, Claude 3, Gemini Ultra, and Llama 4. The model achieved 97.2% accuracy on invoice data extraction and 94.8% on table interpretation tasks. XAI noted that Grok 4.5 demonstrated superior contextual understanding for handling complex invoice layouts and multi-language documents.
Source: xAI Grok (GNews) —
original