Cohere's Parse 5, a 2.3 billion parameter model that converts PDFs, slides, and images to structured Markdown in a single pass, scores 79.2 on ParseBench — below GPT-5.5's 84.4 but ahead of LlamaParse — at a cost of one dollar fifty per thousand pages. The GPT-5.5 equivalent is approximately ninety dollars per thousand pages. For a financial services organisation processing seven hundred and fifty million documents annually, the cost differential exceeds ninety-eight per cent. Document parsing is the most widely adopted enterprise AI use case, deployed by sixty-two per cent of enterprise AI programmes. Parse 5 prices it as a commodity, not as a premium service — with direct implications for the RAG pipelines that underpin the majority of enterprise AI deployments and for the revenue models of cloud vendors whose document AI products carry substantially higher margins.
LLMs
Google Releases Gemini 4 Argon — Tops Frontier Benchmarks and Targets Autonomous Cybersecurity
30 September: Google's Gemini 4 Argon leads the Vals AI benchmark index over OpenAI GPT-6 Astra and …