Bring Your Own AI Key — Now for Text, Images, Video, and Chat
Your AI, your keys, your costs
When we launched BYOK (Bring Your Own Key) last week, it covered text generation — captions and on-design text. Today we're expanding it to cover everything: image generation, video generation, and Max chat.
If you bring your own key, we don't charge credits for those features. You pay your provider directly and keep full control over costs.
What's covered now
| Feature | Key type | What it powers |
|---|---|---|
| Text generation | OpenRouter | Captions, on-design text, post generation |
| Max chat | OpenRouter | All conversations with Max (your AI assistant) |
| Image generation | Google AI (Gemini) | AI-generated images for posts |
| Video generation | Google AI (Gemini) | AI-generated videos |
Two keys, full coverage
You need at most two keys to cover everything:
- OpenRouter key — for text generation and Max chat. Get one at openrouter.ai/keys. Pick from 25+ models across OpenAI, Google, Anthropic, xAI, Meta, Mistral, DeepSeek, and Qwen.
- Google AI key — for image and video generation. Get one at aistudio.google.com/apikey.
Both are optional and independent. Use one, both, or neither.
How to set up
Head to Organization Settings and look for the "Bring your own API key" sections:
- Paste your key — We verify it works before saving.
- Pick your model (OpenRouter only) — Prices are shown in the dropdown so you know what you're spending.
- Done — Your next actions will use your key. No credits charged for features covered by your key.
What about credits?
When your key is active, we skip credit charges for the features it covers. You pay your provider directly at their rates. If you remove your key, you go back to using our default models and credits are charged as usual.
What if I don't set this up?
Nothing changes. We default to solid models for each feature and charge credits normally. BYOK is entirely opt-in for users who want more control over model choice and costs.
Models available today
Through OpenRouter, we support models from seven providers, ranging from $0.10 to $9.00 per million tokens:
- OpenAI: GPT-5 Nano, GPT-4.1 Nano/Mini/Full, GPT-5.4 Mini, o3, o4 Mini
- Google: Gemini 2.5 Flash Lite, Gemini 3 Flash, Gemini 3.1 Flash Lite, Gemini 3.1 Pro
- Anthropic: Claude 4.5 Haiku, Claude Sonnet 4.6, Claude Opus 4.6
- xAI: Grok 4.1 Fast, Grok 4.20, Grok 4
- Meta: Llama 4 Scout, Llama 4 Maverick
- Mistral: Small 3.2, Medium 3.1, Large 3
- DeepSeek: V3, R1
- Qwen: Qwen3 32B, Qwen3 235B
We'll keep adding models as new ones ship.
Take the next step with Marky
Building a successful business in today's digital world requires the right tools and strategies. Marky simplifies the social media process, allowing you to focus on delivering exceptional results.
Ready to streamline your process and grow your business? Visit our landing page to learn more about how Marky can transform your social media strategy today.
Keep reading
Let's do this
Join millions and start posting on-brand content with Marky today.
