Azure Foundry Hosted Agents vs TurboOCR

Microsoft Azure's Foundry Agent Service now offers Hosted Agents in public preview — per-session isolated compute sandboxes purpose-built for running AI agents at scale. Each session gets its own container with a persistent filesystem, internet access (optional), and a Python environment pre-loaded with common agent dependencies. Sessions spin up in seconds and terminate — and stop billing — the moment the agent task completes. The design is framework-agnostic: it officially supports LangGraph, OpenAI Agents SDK, Claude Agent SDK, and Microsoft's own Agent Framework, with others planned. This removes one of the most awkward parts of deploying agents in production: figuring out where they actually run. The persistent filesystem per session means agents can read and write files across their task without external storage configuration. Pricing is $0.0994/vCPU-hour and $0.0118/GiB-hour — competitive with Lambda/Cloud Run for bursty workloads. The service is available in six Azure regions at launch. For enterprises already invested in Azure, this is a compelling "we just figured out the infra" moment. Independent developers can also use it without an enterprise agreement.

TurboOCR is a C++20 OCR server that uses CUDA and TensorRT to process documents at speeds that make Python-based OCR look like a fax machine. The headline number: 270 images per second on FUNSD form datasets with approximately 11ms single-request latency — roughly 50x faster than PaddleOCR's standard Python implementation. It uses PP-OCRv5 models (the same underlying tech as PaddleOCR) but squeezes them through TensorRT FP16 optimization for GPU inference. The server exposes both HTTP and gRPC interfaces from a single binary and handles PDFs natively with four extraction strategies: pure OCR, native text layer extraction, hybrid verification mode, and a "best of both" fallback chain. PP-DocLayoutV3 handles layout detection across 25 document region classes — useful for structured documents where you need to know that a bounding box is a table cell vs. a header vs. a figure caption. A Prometheus metrics endpoint tracks throughput, latency, and GPU memory in real time. Deployment is Docker-first: TensorRT engine compilation happens automatically on first startup. The catch is it requires Linux with an NVIDIA Turing GPU (RTX 20-series minimum) and driver 595+, so it's not a laptop tool. But for enterprise document automation — invoices, forms, medical records — the throughput-to-cost ratio is hard to beat.