Model pricing & capability board for DeepSeek Harness: ~7,250 models across 213 providers, honest lowest-listed comparison, session costs priced by the route you actually called, and cache-leak attribution. Local-first, no server.
Grounded web search for DeepSeek Harness on any OpenRouter-compatible gateway: the built-in web_search tool runs on the same endpoint and key as your chat models.
Cut LLM costs: route routine tasks to cheap subagent models (DeepSeek Harness plugin)
Local OCR for DeepSeek Harness: Apple Vision on macOS, no install, no network, no capability claims
A deterministic /review command for DeepSeek Harness: a reviewer subagent pinned to a subscription-backed Codex model, with a settings card, one reviewer session per chat, and /review reset to forget it.
