"Understanding Inference Models" matching MCP servers:

Multi-model vision understanding MCP server that provides unified image analysis for AI assistants without native vision, supporting models like GLM-4.6V, DeepSeek-OCR, Qwen3-VL-Flash, and more.

Enables non-multimodal models to see images by providing MCP tools for image understanding and OCR, backed by any OpenAI-compatible vision model.

Turns product ideas into rendered UI screen mockups and a buildable component manifest, so downstream agents can develop real screens from the generated designs.

Enables any MCP client to perform image understanding and OCR via any OpenAI-compatible vision-language model. Supports local, private inference without images leaving the machine.

Enables local video inspection and offline transcription with timestamped JSON and Markdown output, enforced filesystem boundaries, and reproducible content-addressed caching.

An MCP server for image understanding via OpenAI-compatible vision models, offering tools for OCR, error screenshot diagnosis, technical diagram reading, data visualization analysis, UI-to-code conversion, and UI diff comparison.

Enables agents to plan on-site AI inference deployments with vendor-neutral hardware sizing, specification-first BOMs, sourcing evidence screening, compliance review, and staged deployment plans.

Exposes an OpenAI-compatible inference endpoint to MCP hosts, enabling them to list available models and run non-streaming chat completions as tools.

Saves the user's exact words as immutable, SHA-256-hashed captures before anything else, then derives events, entities, context cards, and causal hypotheses — every fact traceable to its verbatim source. Local-first personal memory for Claude Code, Codex, and any MCP client.

Enables incremental model scoring over an append-only event log, with a tamper-evident, hash-chained audit trail so every decision can be reviewed and independently re-verified.

The server facilitates natural language interactions for exploring and understanding codebases, providing insights into data models and system architecture using a cost-effective, simple setup with support for existing Claude Pro subscriptions.

Enables local AI inference using Ollama and open-weight models, designed for offline operation in low-connectivity environments.

Enables integration of local LLM capabilities with MCP-compatible clients like Claude Desktop, Continue.dev, and Cline. Provides tools for processing text prompts through local language models using a customizable inference function.

An MCP server that wraps the @apitomy/data-models library for querying, validating, and editing OpenAPI and AsyncAPI documents.

Enables managing Onde Inference accounts and model catalog operations through MCP tools such as login, app management, model registration, and assignment. Returns structured JSON over stdio for use with any MCP client.

Provides an MCP-compatible profiler for measuring edge and on-device LLM inference latency, including TTFT, TPOT, tokens-per-second throughput, and P50/P90/P99 jitter distributions. It enables agents and developers to run these telemetry benchmarks from MCP clients such as Claude Desktop, Cursor, and Windsurf.

Provides context management, task classification, and routing tools for small models (1B-12B) in production agentic workflows.

Enables image understanding, OCR, multi-image comparison, and AI image generation through pluggable models, supporting tasks like visual analysis, text extraction, and image synthesis.

Adds image understanding to any MCP-compatible AI client via free Zhipu vision models. Enables agents to analyze local images or URL images by answering natural language questions.

Enables agents to inspect GPU status, size KV cache and vLLM/llama.cpp deployments, read live backend metrics, and run benchmarks over MCP.