AI Integration

Butterbase includes a built-in AI model gateway that lets your app call large language models through an OpenAI-compatible API. You can use the platform’s shared key or bring your own key (BYOK) for direct billing with the model provider.

How it works

Your app sends chat completion or embedding requests to Butterbase, which proxies them to the model provider (via OpenRouter). Usage cost is tracked automatically and counted against your plan’s AI credits allowance.

Chat completions

POST /v1/{app_id}/chat/completions
Authorization: Bearer {token}

{
  "model": "anthropic/claude-3.5-sonnet",
  "messages": [
    { "role": "system", "content": "You are a helpful assistant." },
    { "role": "user", "content": "What is Butterbase?" }
  ],
  "max_tokens": 500,
  "temperature": 0.7
}

Streaming: Set "stream": true to receive server-sent events.

Embeddings

Generate vector embeddings for semantic search, clustering, and other ML tasks:

POST /v1/{app_id}/embeddings
Authorization: Bearer {token}

{
  "model": "openai/text-embedding-3-small",
  "input": "What is Butterbase?",
  "encoding_format": "float"
}

Batch input: Pass an array of strings:

{
  "model": "openai/text-embedding-3-small",
  "input": ["first text", "second text", "third text"]
}

Available embedding models

Model	ID	Dimensions
Text Embedding 3 Small	openai/text-embedding-3-small	1536
Text Embedding 3 Large	openai/text-embedding-3-large	3072
Text Embedding Ada 002	openai/text-embedding-ada-002	1536

Available models

Butterbase supports any model available through OpenRouter:

Model	ID
Claude 3.5 Sonnet	anthropic/claude-3.5-sonnet
Claude 3 Opus	anthropic/claude-3-opus
Claude 3 Haiku	anthropic/claude-3-haiku
GPT-4 Turbo	openai/gpt-4-turbo
GPT-4	openai/gpt-4
GPT-3.5 Turbo	openai/gpt-3.5-turbo
Llama 3.1 70B	meta-llama/llama-3.1-70b-instruct
Llama 3.1 8B	meta-llama/llama-3.1-8b-instruct

The full list is available at openrouter.ai/models.

AI configuration

Configure AI settings per app:

PUT /v1/{app_id}/ai/config

{
  "defaultModel": "anthropic/claude-3.5-sonnet",
  "byokKey": "sk-or-...",
  "maxTokensPerRequest": 4096,
  "allowedModels": ["anthropic/claude-3.5-sonnet", "anthropic/claude-3-haiku"]
}

Setting	Description
defaultModel	Model used when none is specified
byokKey	Your own OpenRouter API key (encrypted at rest)
maxTokensPerRequest	Maximum tokens per request (1–100,000)
allowedModels	Restrict which models can be used

Bring Your Own Key (BYOK)

By default, requests use the platform’s shared key and count against your plan’s AI credits. With BYOK:

• Requests are billed directly to your OpenRouter account
• Usage is not counted against your Butterbase quota
• Your key is encrypted at rest

Usage tracking

GET /v1/{app_id}/ai/usage?startDate=2026-01-01&endDate=2026-01-31

Returns total tokens, cost, and breakdown by model.

Using AI in serverless functions

The runtime auto-injects BUTTERBASE_APP_ID and BUTTERBASE_API_URL — you only need to supply your API key via envVars:

export default async function handler(req: Request, ctx: any): Promise<Response> {
  const { BUTTERBASE_APP_ID, BUTTERBASE_API_URL, BUTTERBASE_API_KEY } = ctx.env;

  const aiResponse = await fetch(`${BUTTERBASE_API_URL}/v1/${BUTTERBASE_APP_ID}/chat/completions`, {
    method: 'POST',
    headers: {
      'Content-Type': 'application/json',
      'Authorization': `Bearer ${BUTTERBASE_API_KEY}`
    },
    body: JSON.stringify({
      model: 'anthropic/claude-3-haiku',
      messages: [{ role: 'user', content: 'Summarize this text: ...' }],
      max_tokens: 200
    })
  });

  const result = await aiResponse.json();
  return new Response(JSON.stringify(result), {
    headers: { 'Content-Type': 'application/json' }
  });
}

AI credits

Plan	AI credits	Resets?
Free	$0.10	No — lifetime allowance
Pro	$10.00/mo (then $0.10/credit overage)	Yes — resets each billing period
Enterprise	Unlimited	—

BYOK usage is not counted against these limits.