BiteBrief API v3.1

Summarize any URL, YouTube video, podcast, or PDF with a single API call. Powered by Gemini 3.1 Pro/Flash with intelligent model routing and automatic fallback.

Key Features

• Gemini-first: Gemini 2.5 Flash (speed) and Pro (quality) with OpenRouter/Claude fallback
• Multimodal: Native audio/video processing without external transcription
• Smart routing: Auto-selects best model by content type, length, and user tier
• Caching: Identical requests return cached results instantly at zero LLM cost
• Cost tracking: Every response includes model, tokens, cost, and latency
• Premium features: Audio Overview, Q&A, Mind Map, Translation
• Agent-ready: x402 protocol for autonomous AI agent payments (USDC on Base)
• Cost Guard: Real-time margin enforcement, minimum 3x markup on all operations

Authentication

All API requests require a Bearer token in the Authorization header. x402 agent endpoints use payment-based auth instead.

Authorization: Bearer YOUR_API_KEY

API keys are provisioned during testing. Contact hello@bitebrief.com for access.

Quickstart

curl -X POST https://api.bitebrief.com/v1/summarize \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://youtube.com/watch?v=dQw4w9WgXcQ",
    "format": "markdown",
    "length": "medium"
  }'

POST /v1/summarize

Summarize any URL. The system auto-detects content type and routes to the optimal model.

Request Body

Response

{
  "id": "sum_a1b2c3d4e5f6",
  "url": "https://example.com/article",
  "type": "webpage",
  "summary": "# Key Points\n\n- Point one...\n- Point two...",
  "format": "markdown",
  "word_count": 280,
  "source_word_count": 5200,
  "title": "Article Title",
  "processing_time_ms": 1840,
  "timestamp": "2026-02-25T18:00:00",
  "model": {
    "name": "gemini-2.5-flash",
    "provider": "gemini",
    "input_tokens": 8200,
    "output_tokens": 480,
    "cost_usd": 0.0010,
    "latency_ms": 1620,
    "routing_reason": "webpage content → gemini-2.5-flash"
  }
}

Supported Content Types

POST /v1/quote

Get a price quote before summarizing. Required when dynamic pricing is enabled. Quotes are valid for 24 hours and single-use.

{ "url": "https://example.com/podcast.mp3" }

{
  "quote_id": "qt_abc123def456",
  "url": "https://example.com/podcast.mp3",
  "credits_required": 60,
  "cost_usd": 1.49,
  "backend_cost": 0.48,
  "margin_percent": 67.8,
  "tier": "long",
  "expires_at": 1740600000
}

POST /v1/batch

Submit up to 100 URLs for batch processing. Each URL is summarized independently.

{
  "urls": [
    "https://example.com/article1",
    "https://example.com/article2"
  ],
  "format": "markdown",
  "length": "short"
}

Audio Overview +5 credits

Generate an executive-style briefing from audio/podcast content. Includes speaker insights, action items, and tone analysis.

{ "content": "transcript text...", "url": "https://..." }

Q&A Follow-up +3 credits/question

Ask follow-up questions about any previously summarized content.

{
  "question": "What were the main arguments?",
  "summary": "Previously generated summary...",
  "content": "Original content (optional)..."
}

Mind Map +5 credits

Extract a hierarchical topic structure as structured JSON.

{ "content": "article or transcript text..." }

Translation +2 credits

Translate any summary to 30+ languages. Use GET /v1/features/languages for the full list.

{ "summary": "Summary text...", "language": "es" }

x402 Agent-to-Agent Payments

AI agents can use BiteBrief without an API key by paying per request via the x402 protocol. Payments in USDC on Base chain.

Wallet: Payments to sovereign wallet on Base chain. Minimum per-request pricing enforced.

Agent Quoting

{ "url": "https://example.com/article" }

{
  "quote_id": "qt_...",
  "cost_usd": 0.15,
  "wallet": "0x43E03211a163A126999393Ab6a6A950FC7fc3dC6",
  "chain": "base",
  "currency": "USDC",
  "expires_at": 1740600000
}

Model Routing

BiteBrief automatically selects the optimal model based on content type, duration, and user tier. You can override with the model parameter.

{
  "models": [
    { "model": "gemini-2.5-flash", "provider": "gemini", "tier": "standard" },
    { "model": "gemini-2.5-pro", "provider": "gemini", "tier": "premium" },
    { "model": "anthropic/claude-sonnet-4", "provider": "openrouter", "tier": "standard" }
  ]
}

Caching

Responses are cached by sha256(url + model + options). Cache hits return instantly at zero LLM cost. TTL: 24 hours.

• Force fresh: Set "nocache": true in the request body
• Cache stats: GET /api/cache
• Clear cache: POST /api/cache/clear (admin)

Rate Limits

Hard rate limits are enforced per API key to prevent abuse and ensure fair usage.

Exceeding limits returns 429 Too Many Requests with a Retry-After header.

Cost Guard

Every request is checked in real time against the Cost Guard system. If the projected margin falls below the minimum threshold (3x cost), the request is blocked.

Anti-Abuse Policy

Designed to be highly profitable. Abuse = automatic suspension.

Metrics

Real-time LLM performance metrics and cache statistics.

{
  "llm": {
    "total_llm_calls": 142,
    "total_llm_cost_usd": 0.2840,
    "models": {
      "gemini/gemini-2.5-flash": {
        "calls": 120, "success_rate_percent": 99.2,
        "avg_cost_usd": 0.0012, "avg_latency_ms": 1200
      }
    }
  },
  "cache": {
    "entries": 89, "hits": 34, "misses": 108,
    "hit_rate_percent": 23.9,
    "estimated_savings_usd": 0.17
  }
}

Health & Status