Use your own API keys, pay providers direct, keep full control
SleekAI talks directly from your WordPress server to OpenAI, Anthropic, Google, or OpenRouter using the API keys you paste into the dashboard. There’s no proxy, no token markup, and no per-message surcharge. Your prompts and conversations stay between your server and the model vendor of your choice.
Per-message pricing makes AI chat expensive at scale
A site that pays $50 a month for the platform plus its own model usage often ends up at $300 once volume picks up, with little control over what part of the bill is model and what part is platform tax.
SleekAI is a WordPress plugin, not a service. You paste your OpenAI, Anthropic, Google, or OpenRouter API key into the settings page once. From then on, every chat turn sends a request directly from your server to the provider you chose.
The model vendor bills you directly for tokens used. SleekAI is paid once for the license and never takes a cut of inference. Your token budget is whatever your provider account is set to, not whatever a credit-bundle reseller decides to limit you at.
Generic platforms argue that bundling tokens simplifies billing. In practice it removes visibility and limits choice. With BYOK you can run a cheap Gemini Flash bot on a high-volume FAQ and an expensive Claude Opus bot on a complex sales flow, pay exactly what each consumes, and switch providers in seconds without the platform telling you which models are ‘supported’ on your plan tier.
Features
Direct provider keys
Paste your OpenAI, Anthropic, Google, or OpenRouter API key into the SleekAI settings. From that moment, every inference request goes straight from your server to the provider. SleekAI never sees your prompts and never proxies the request through any external server.
Provider portability
Change the model dropdown on a bot to flip between providers. Same prompt, same conversation state, same WordPress variables. If a provider raises prices or releases a worse model, you switch in 30 seconds without rebuilding anything or migrating data.
Answers from your data
Map post titles, content, post meta, taxonomies or a WP_Query into the instruction, and the bot answers from what your site actually publishes.
Why teams choose BYOK over bundled inference
Predictable per-token billing
Finance teams prefer the direct provider invoice because it’s auditable. Token usage appears in the OpenAI or Anthropic console with no platform abstraction layer. Variances between months are explainable and tied to actual model consumption.
Volume scaling without surcharge
Scaling from 10k chats to 1 million chats per month doesn’t trigger platform-tier upgrades. The model bill scales linearly with usage but the SleekAI license stays at the original one-time price. Volume customers save dramatically.
Repeat questions
Answer the questions visitors ask every day from the pages and posts where you already answer them.
How BYOK works inside SleekAI
Paste your provider keys
Open the SleekAI providers panel and paste your API keys for OpenAI, Anthropic, Google, OpenRouter, or any custom OpenAI-compatible endpoint.
Pick a model per bot
Each chatbot has a model dropdown. Choose any model whose provider you’ve configured. The bot uses that exact model for every turn. Different bots on the same site can use different providers and different models without any cross-bot interference.
Inference goes direct
When a user sends a message, SleekAI builds the request and posts it directly from your WordPress server to the chosen model vendor’s API. The vendor processes the inference and returns a reply. No external proxy ever sees the content.
Pay providers, not platforms
The model vendor invoices you for actual token usage at their published rates. SleekAI is paid once for the plugin license and never takes a percentage of inference. Volume scaling works linearly with token cost, with no platform-tier upgrades, no overage fees, and no per-message surcharge.
Why BYOK is the right architecture for chat
Hosted chat platforms made sense when frontier APIs were hard to integrate and inference was expensive enough that bundling felt like a discount. Neither is true anymore. OpenAI, Anthropic, and Google APIs are straightforward to call from any server with an HTTP client, and direct provider pricing has dropped to fractions of a cent per typical chat turn.
The economic case for paying a platform 2-5x markup on top of the model itself has eroded. What remains is the convenience argument, and that argument falls apart at any meaningful scale.
The other dimension is control. With BYOK, the model vendor relationship is yours directly. Their data handling, retention, and abuse policies apply to you on their terms, with no platform intermediary adding a fourth party to the data flow.
If a vendor releases a better model or cuts prices on an existing one, you switch immediately by changing a dropdown. The whole stack collapses to two parties, your server and the model vendor, which is the simplest architecture that can possibly work.
Frequently asked questions
Which providers are supported?
OpenAI (GPT-4o, GPT-4o mini, o1, o3 family), Anthropic (Claude 3.5 Sonnet, Haiku, Opus, Claude 3.7), Google (Gemini 1.5 Pro, Flash, 2.0 family), and OpenRouter (which itself spans 200+ models from many providers). Self-hosted endpoints that speak the OpenAI-compatible API also work via the custom endpoint option.
Does SleekAI proxy any requests?
No. Every inference request goes directly from your WordPress server to the model vendor’s API using your key. The plugin runs entirely on your hosting infrastructure.
What does SleekAI cost?
After the license you only pay the model vendors for actual usage, billed directly through their consoles at whatever rate they publish.
Can I use a self-hosted model?
Yes. The custom endpoint option accepts any OpenAI-compatible API URL. Ollama with the OpenAI compatibility layer, vLLM, llama.cpp server, and several other inference servers all work. Point SleekAI at your endpoint URL, optionally include an auth header, and bots can use your self-hosted model with the same configuration as any cloud provider.
Which AI providers does it support?
OpenAI, Anthropic, Google and OpenRouter, plus custom OpenAI-compatible endpoints. You bring your own API key and the provider bills you directly.
Where does the chatbot get its answers?
From the WordPress data you map into its instruction: post titles and content, post meta, ACF and Meta Box fields, taxonomies, terms, users and user meta, or a WP_Query. Values resolve on every request.
Can it act on a visitor’s behalf?
No. A chatbot answers questions. It does not submit forms, create bookings, open tickets or send email; put the links for those steps in its instruction so it can point visitors to them.
Can I control where the chatbot appears?
Yes. Display conditions cover post type, post, post status, post template, taxonomy, term, user role, logged-in state and WordPress conditional tags such as front page, archive, search and 404.
Pricing
Explore our flexible licensing options tailored to your needs. Upgrade your license anytime to access more features, or opt for a lifetime license for ongoing value, including lifetime updates and lifetime support. Our hassle-free upgrade process ensures that our platform can grow with you, starting from whichever plan you choose.
…or get the Bundle Deal
and save €250
The Bundle (unlimited sites)
Pay once, own it forever
Elevate your WordPress site with our exclusive plugin bundle that includes all of our premium plugins in one package. Enjoy lifetime updates and lifetime support. Save significantly compared to buying plugins individually.
What’s included
SleekAI
SleekByte
SleekMotion
SleekPixel
SleekView
SleekRank
SleekTrack