DocJev

An open-source tax-document page classifier built with TypeSafe Jev. It extracts text from PDF pages, then uses decision-model requests to classify each page by form and page kind, returning probability distributions, a form identifier, confidence scores, and a gate indicating whether the result meets the configured confidence threshold. Its generated JSON criteria file describes IRS forms using data extracted from IRS form PDFs; a second request handles schedules for selected corporate and foreign forms. The package includes PDF helpers based on Poppler, supports text-layer input, and requires a TypeSafe API key for the Jev backend. It does not train or host a model, is limited to English federal-form classification, requires OCR for scanned pages, and is distributed under Apache-2.0 with separate licensing for IRS-derived data.

Docling is an open-source document-processing toolkit from the Docling project for preparing files for generative-AI applications. It parses formats including PDF, DOCX, PPTX, XLSX, HTML, EPUB, images, plain text, audio, and video, and represents the results in a unified DoclingDocument structure. For PDFs, it analyzes page layout, reading order, headings, tables, code, formulas, images, and scanned content through OCR. Audio and video inputs can be processed with automatic speech-recognition models; video parsing produces an ASR transcript and representative keyframes. Parsed content can be exported as Markdown, HTML, WebVTT, DocLang, DocTags, or lossless JSON. Docling supports local and air-gapped execution, a command-line interface, Python usage, an API server named docling-serve, and an MCP server for connecting agents. The project also provides integrations for LangChain, LlamaIndex, CrewAI, and Haystack. It is installable with pip and runs on macOS, Linux, and Windows on x86_64 and arm64 systems.

Jevmind is a Python decision layer for software agents. It turns agent decisions into typed answers such as yes/no results, choices, and scores with confidence values, then applies a code-defined gate: uncertain answers escalate, trusted positive answers may act, and trusted negative answers are held. Its decisions are recorded in an append-only, hash-chained ledger, graded against later outcomes, and optionally recalibrated with Platt scaling. It includes nine skills for compacting tool output, navigating repositories, reviewing diffs, routing tasks to model tiers, checking completion claims, curating training data, walking linked markdown vaults, guarding commands, and running a structured-state control-loop demo. The default local brain uses readable rules, reflexes, and BM25 retrieval without a language model; an optional Jev HTTP brain can answer the same typed questions, while replay mode uses recorded answers. Jevmind runs as a command-line tool and MCP server, includes a Claude Code hook for shell-command screening, supports Python 3.10+, has no runtime dependencies, and is distributed under the MIT license.

Simple Jev is an open-model classifier and scoring server from Featherless AI. It accepts shared context or chat messages plus questions, then reads the model's next-token logits for predefined answer labels instead of generating a prose or JSON completion. The server constructs JSON responses for choice classification, ordered rubric scores, and truth or support judgments, including confidence and label distributions. Its Hugging Face Transformers implementation identifies the exact common token prefix across questions, evaluates that prefix once, reuses its KV cache for question-specific suffixes, and batches the suffixes before scoring the allowed answer tokens. The response-scoring code normalizes the selected logits; shared context is counted once for usage purposes, and no output tokens are generated. The API provides non-streaming /v1/classifier and /v1/systemone endpoints, with local execution supported through Python, PyTorch, and Transformers. A public demo API is available, while production deployments can use Featherless paid plans or run the server themselves. The repository also includes RFDT, which fine-tunes models directly on allowed answer-token logits using the same prompt structure.

jeff is a self-hosted, MIT-licensed drop-in implementation of TypeSafe's jev System One API, powered by the GLiFormer encoder model. It accepts classification questions through a compatible SDK or the POST /v1/systemone endpoint and returns choice selections, ordered scores, and noul yes/no probabilities. The server supports CUDA, Apple MPS, CPU, and ONNX Runtime deployments, with batching, bearer-key authentication, rate limits, request limits, health and statistics endpoints, and deployment configurations for Modal. Its repository reports lower hosting costs than jev but weaker results on reasoning-heavy tasks.