---
title: "Guides"
canonical: https://pentestllm.ai/guides
generated: 2026-11-18T03:25:53.510Z
source: https://pentestllm.ai/guides
base_url: https://api.pentestllm.ai/v1
---

# Guides

- [Open WebUI add model: security builds on one key](https://pentestllm.ai/guides/open-webui) (steps checked 2026-10-25): How to add a model to Open WebUI from the security catalog: a single OpenAI connection, a picker trimmed on the key, one preset per build with a plain system prompt, and the context window each build gives a long engagement chat.
- [LibreChat custom endpoint for security models](https://pentestllm.ai/guides/librechat) (steps checked 2026-11-06): Point a LibreChat custom endpoint at the security catalog: the endpoints.custom block in librechat.yaml, the key variable in .env, the /v1 base URL, a models.default list and a cheap titleModel. Plus the Docker mount, what a chat costs and how to read each error.
- [Ollama uncensored model list for security work, and the API builds that replace them](https://pentestllm.ai/guides/ollama-uncensored-models) (steps checked 2026-10-27): Each ollama uncensored model tag on the library, as read on the date below, beside the same-family or newer abliterated build on one prepaid key: how big the download is, the live per-million price, what 100 triage messages cost, and the two fields that move Open WebUI or AnythingLLM off the local GPU.
- [A LiteLLM OpenAI-compatible endpoint for security models, with a fallback chain](https://pentestllm.ai/guides/litellm) (steps checked 2026-10-25): LiteLLM against this endpoint: the openai/ prefix, a model id and api_base, in the Python SDK and the proxy config.yaml, with three builds chained as fallbacks.
- [n8n OpenAI-compatible API: a security model behind the AI Agent](https://pentestllm.ai/guides/n8n) (steps checked 2026-10-25): Wire an n8n OpenAI credential to the security catalog: a custom Base URL, the Chat Model sub-node hung off an AI Agent, the id typed in ID mode, the Responses switch turned off, and what a thousand triage runs cost.
- [An uncensored Discord bot for a security team, in one Python file](https://pentestllm.ai/guides/discord-bot) (steps checked 2026-10-25): Run an uncensored Discord bot on your team server: one discord.py script, the OpenAI SDK on this API, four variables, and answers only where you allow them.
- [Change the OpenAI SDK base URL to an uncensored security model](https://pentestllm.ai/guides/sdks) (steps checked 2026-10-26): Set OPENAI_BASE_URL plus OPENAI_API_KEY and your OpenAI SDK code calls an uncensored build: Python, TypeScript, curl, the AI SDK, streaming, JSON, errors, cost.
- [Point the AI SDK OpenAI-compatible provider at a pentest model](https://pentestllm.ai/guides/ai-sdk) (steps checked 2026-11-02): createOpenAICompatible once, then generateText, streamText and tool() against an uncensored build: tool support, usage fields, a Next.js route, errors, cost.
- [Claude Code on an uncensored model for security work](https://pentestllm.ai/guides/claude-code) (steps checked 2026-10-26): Run a security-tuned or uncensored build behind Claude Code with claude-code-router: a provider block on this API, two shell variables, one curl that proves it.
- [Cursor custom API key: a security-tuned model in the chat panel](https://pentestllm.ai/guides/cursor) (steps checked 2026-10-26): Enter your key in Cursor’s OpenAI API Key field, swap the base URL, add a security or uncensored id to the picker. Works in Ask and Agent; errors and fixes.
- [Cline OpenAI compatible provider: run a security model in VS Code](https://pentestllm.ai/guides/cline) (steps checked 2026-11-02): Switch Cline’s API Provider to OpenAI Compatible, enter the base URL, the key and a model id, then type the Model Configuration values the build needs. Split Plan and Act across two models, read the cost of an agent turn and fix the errors Cline reports.
- [Roo Code OpenAI compatible provider: a security model per mode](https://pentestllm.ai/guides/roo-code) (steps checked 2026-11-02): Give Roo Code’s OpenAI Compatible provider the Base URL, the API Key and a Model id, then copy that build’s Model Configuration row, which this page reads from the catalog at render. A profile for each mode, what an agent turn costs and the errors Roo reports.
- [Kilo Code OpenAI compatible provider: security models in the extension and the CLI](https://pentestllm.ai/guides/kilo-code) (steps checked 2026-11-02): Two ways to add the API to Kilo Code: the Custom provider dialog under Providers, or a kilo.jsonc block with the base URL, the key and each model id with its cost and limits. Extension and CLI, what an agent turn costs and the errors.
- [Continue.dev on a security model that answers](https://pentestllm.ai/guides/continue) (steps checked 2026-11-02): Point Continue at this API with a single models entry in config.yaml: provider openai, apiBase, a secret for the key and tool_use so Agent mode works. Covers VS Code, JetBrains and the cn CLI, what an Agent turn costs, and the errors you may hit.
- [Aider on a security model: the OpenAI compatible API setup](https://pentestllm.ai/guides/aider) (steps checked 2026-11-02): Point aider at the API with two environment variables or .aider.conf.yml, prefix the model id with openai/, and silence the unknown-model line with a metadata file. Weak model, architect mode, what an edit turn costs and the errors you will meet.
- [Zed OpenAI-compatible provider: a security-tuned model in the Agent Panel](https://pentestllm.ai/guides/zed) (steps checked 2026-11-02): The openai_compatible block for Zed’s settings.json: api_url, security or uncensored ids with max_tokens and capabilities, the key in keychain or env var.
- [LangChain on an OpenAI compatible API, for security work](https://pentestllm.ai/guides/langchain) (steps checked 2026-11-02): Python ChatOpenAI(base_url, api_key) or JavaScript configuration.baseURL, an uncensored build id, and the create_agent tool loop on it: usage, cost, errors.
- [OpenCode on a security model that answers](https://pentestllm.ai/guides/opencode) (steps checked 2026-11-02): Wire OpenCode to this API with a single provider entry in opencode.json naming the openai-compatible package, the base URL, an environment variable for the key and the ids /models should list. Then what an agent turn costs and how to read each error.
- [Hermes Agent on a security model that answers](https://pentestllm.ai/guides/hermes-agent) (steps checked 2026-11-06): Add this API to Hermes Agent as one providers entry in ~/.hermes/config.yaml: the api URL, a key_env name, the chat_completions transport and your model ids. The key stays in .env, /model switches builds mid-session, and the page prices an agent turn and decodes each error.
- [OpenClaw on a security model that answers](https://pentestllm.ai/guides/openclaw) (steps checked 2026-11-06): Give OpenClaw this API through one models.providers entry in ~/.openclaw/openclaw.json: baseUrl, a key pulled from the environment, the openai-completions adapter and a models array of security and abliterated builds. Then the default model, what a session costs and the error fixes.
- [Codex CLI custom provider for security-tuned models](https://pentestllm.ai/guides/codex-cli) (steps checked 2026-11-06): A [model_providers] table in ~/.codex/config.toml: base_url, env_key, wire_api responses, model on a security build. A profile per model; the /model menu trap.
- [Set up an uncensored coding AI for security work in Claude Code, Cursor and Open WebUI](https://pentestllm.ai/guides/uncensored-coding-ai) (steps checked 2026-10-27): An uncensored coding AI inside the agent a tester already runs: the code builds with the refusal trained out, wired into Claude Code via a router, Cursor and Open WebUI, one agent day priced from the live catalog, and the line the acceptable-use page draws.
- [The uncensored image API and the video route, for report graphics](https://pentestllm.ai/guides/media) (steps checked 2026-10-26): Two routes for the image and video models: the OpenAI image call and a generations endpoint you poll. Requests, price-setting fields, a poll loop, refunds.
