dsh-our-free-model

在 dsh 里装上这个插件即可,无需登录、注册或填 API Key,就能使用包括 DeepSeek V4.1 Flash、Kimi K3 在内的前沿模型——完全免费,不限量。 All you do is install this p

What is dsh-our-free-model?

在 dsh 里装上这个插件即可,无需登录、注册或填 API Key,就能使用包括 DeepSeek V4.1 Flash、Kimi K3 在内的前沿模型——完全免费,不限量。 All you do is install this plugin in dsh: no login, no sign-up, no API key — the frontier models are just there, DeepSeek V4.1 Flash and Kimi K3 among them. Completely free, with no usage cap.

Overview of dsh-our-free-model

What it is

This is a plugin for dsh that provides free access to frontier large language models such as DeepSeek V4.1 Flash and Kimi K3. No login, registration, or API key is required. The plugin is a pure add-on that does not modify the kernel, requires no build steps, and has zero dependencies.

Key features

  • No account, configuration, or API key needed to start using models immediately.
  • Model list is refreshed from upstream (OpenCode’s Zen gateway and Kilo AI’s public gateway) on each poll; no static snapshot is kept.
  • Selection dropdown only broadcasts models that are currently routable; models explicitly rejected by the gateway are hidden with the reason noted.
  • Built-in announcement center with real-time push via a remotely editable JSON file.
  • In-app upgrade with SHA-256 checksum, atomic replacement, and rollback on failure.
  • Hot-reload of code changes; supports file-watch auto-reload.
  • Thinking intensity (Light / Balanced / Deep) genuinely sets an output-token budget rather than using a prompt trick.
  • Usage dashboard stored entirely locally (token heatmap, total curves, speed and first-token latency).
  • OpenAI-compatible local forwarding port for other tools.
  • EAC channel (desktop only) unlocked via GitHub authentication and repository starring.
  • Kilo channel built-in free pool via Kilo AI’s public gateway; no account needed.
  • Aggregation channel that lets you plug in 76 third-party providers (NVIDIA NIM, OpenRouter, Groq, Cerebras, Gemini, etc.) with your own API key stored locally.
  • Token-saving features: tool output compression (RTK), prompt slimming profiles, and five local-answer modes.
  • Routing / fallback chains that compose models into named combos with fallback or round-robin strategies.
  • Native Anthropic-compatible /v1/messages endpoint on the local gateway.
  • Thirteen pre-integrated free proxy channels (CodeArts, CodeBuddy/WorkBuddy, LobsterAI, Qoder, TRAE, Cline, Loomy, Raccoon, MiniMax Code, ZCode, Gemini) plus the Kilo free lane and the original anonymous free lane.

Use cases

Accessing frontier LLMs for free without managing accounts or API keys. Integrating multiple model providers (including local free pools and key-based services) into a single interface. Developing or testing applications that require an OpenAI- or Anthropic-compatible local endpoint with thinking support. Running a dashboard to monitor token usage, latency, and model performance. Using with dsh in terminal-only compositions (dsh-tui) that lack a web server.

Getting started

See the repository README at https://github.com/Ebony-Vinyl/dsh-our-free-model for installation instructions.

License and pricing

License: MIT. Pricing: open source and free for use with no usage cap.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/ebony-vinyldsh-our-free-model.svg)](https://yunodb.com/tools/ebony-vinyldsh-our-free-model?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/ebony-vinyldsh-our-free-model?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/ebony-vinyldsh-our-free-model.svg" alt="Featured on YunoDB" height="20"></a>