Awesome BYOK Apps 
A curated list of apps you can use with your own AI API key.
BYOK (Bring Your Own Key) apps let you plug in your own OpenAI,
Anthropic, Google, OpenRouter, or other provider key instead of paying
a subscription markup. You pay the provider directly for tokens, and
your data stays out of an intermediate SaaS.
Contents
Chat UIs
Multi-provider chat clients — paste your provider key and go.
- AnythingLLM — Document chat, workspaces, and agents; desktop and server.
- BetterChatGPT — Lightweight static web UI for OpenAI.
- big-AGI — Multi-model workspace with personas, tools, and Beam.
- Chatbox — Cross-platform desktop and mobile chat client.
- Jan — Offline-first desktop app; connects cloud providers or runs local models.
- LibreChat — Multi-provider ChatGPT clone with agents, MCP, and RAG.
- LobeChat — Polished multi-provider chat with plugins and MCP support.
- NextChat — Cross-platform (web and Tauri) ChatGPT clone.
- Open WebUI — Self-hosted platform for Ollama and OpenAI with RAG.
- SillyTavern — Power-user roleplay frontend with per-user secrets.
- TypingMind — Polished web client with AES-encrypted local storage.
Code assistants
- Aider — Terminal AI pair-programmer integrated with git.
- Cline — Autonomous coding agent for VS Code.
- Continue — VS Code and JetBrains assistant with custom providers.
- Kilo Code — Agentic VS Code and CLI; fork of Cline and Roo.
- Open Interpreter — Natural-language shell that runs code locally.
- Roo Code — Multi-mode agent team; fork of Cline.
- Tabby — Self-hosted coding assistant you can point at any provider.
- Zed — Rust GPU-accelerated editor with an AI Agent Panel.
IDE and terminal
Built-in BYOK in commercial tools.
- Cursor — AI code editor; add OpenAI, Anthropic, Google, or custom keys.
- JetBrains AI — Plug Anthropic or OpenAI keys into JetBrains IDEs.
- VS Code — Language Model Chat Provider API for custom providers.
- Warp — AI terminal with locally stored BYOK.
- Zed Agent Panel — Supply Anthropic, OpenAI, Google, or Ollama keys.
Writing and research
- Glarity — Browser extension that summarizes search results and YouTube.
- GPT Researcher — Autonomous research agent producing long-form reports.
- Khoj — Personal AI across browser, desktop, and plugins.
- Novel — Notion-style editor with AI commands.
- OpenGrammar — Open-source Grammarly alternative.
- Perplexica — Open-source Perplexity-style search.
Translation
- NextAI Translator — Extension and desktop translator across many providers.
- Read Frog — Immersive-translation browser extension with 20+ providers.
Agents and workflow builders
- CrewAI — Role-based multi-agent framework.
- Flowise — Visual low-code agent builder with encrypted credentials.
- Langflow — Visual AI builder with provider setup and global variables.
- n8n — Workflow automation with node-level LLM provider credentials.
- SuperAGI — Self-hosted autonomous agent framework.
Image, audio, and voice
- ComfyUI — Node-based image pipeline; external provider nodes accept API keys.
- Fabric — CLI for patterned prompt workflows across providers.
- LM Studio — Local and cloud model client with provider key fields.
- Open NotebookLM — Research and podcast generation with your own STT/TTS keys.
Gateways and proxies
Run your own BYOK-enabled backend, or use a hosted aggregator.
- Cloudflare AI Gateway — Store provider keys and route at the edge.
- Helicone — Observability proxy with BYOK for 100+ providers.
- LiteLLM — Self-hosted proxy unifying 100+ LLM providers.
- OpenRouter — Hosted aggregator; one key for hundreds of models.
- Portkey — Managed or self-hosted gateway with budgets and routing.
- Vercel AI Gateway — Team-scoped managed gateway with BYOK support.
Providers
Where to get API keys.
- Anthropic — Claude; create keys in Console → Settings → API Keys.
- Cerebras — Ultra-fast inference for Llama and Qwen.
- Cohere — Command, Embed, and Rerank models.
- DeepInfra — Hosted open-weight inference.
- DeepSeek — DeepSeek-V3 and R1 reasoning models.
- ElevenLabs — Text-to-speech and voice cloning.
- Fireworks AI — Fast open-model hosting.
- Google AI Studio — Gemini family keys with a free tier.
- Groq — Free-tier LPU inference for open models.
- Hugging Face — Inference API across thousands of hosted models.
- Mistral AI — Mistral and Codestral models, EU-hosted.
- OpenAI — GPT, DALL·E, Whisper, and TTS.
- OpenRouter — One key for hundreds of models across providers.
- Perplexity — Sonar search-grounded models.
- Replicate — Run open-source models via HTTP.
- Together AI — Hosted open-weights inference.
- xAI — Grok models.
Contributing
Contributions welcome — see CONTRIBUTING.md.
Rules:
- Only include apps or providers that are live and working.
- Entries follow
- [Name](link) — Short description.
- Alphabetical order within each section (case-insensitive).
- No duplicates; merge close entries.
License
To the extent possible under law, the contributors have waived all
copyright and related or neighboring rights to this work.