Provider Routing
Smart failover and traffic distribution across providers with unified model aliases.
Your coding agents deserve better than manual API key juggling. MAVIS handles the plumbing so you can focus on what ships.
Smart failover and traffic distribution across providers with unified model aliases.
One API key and endpoint for every supported model — no per-provider credentials to manage.
Track latency, volume, and operational metrics for every request path.
Production-ready gateway patterns for retries, fallback, and controlled rollouts.
Works with every agentic coding tool
Every plan includes access to all providers — Claude, GPT, Gemini, and more — through a single API endpoint. Budget refreshes automatically each window.
Matches Claude Pro — 25% less
3× Claude Pro capacity
5× Claude Max capacity, save 25%
Rivals Claude Max 20x — save $50/mo
Need custom budgets, dedicated infrastructure, SLA guarantees, or volume pricing? We'll build a plan that fits.
* Prompt estimates based on real-world usage data from Anthropic. Actual counts vary by model and task complexity.
Claude comparison: Pro ~500.000₫/mo, Max 5x ~2.500.000₫/mo, Max 20x ~5.000.000₫/mo.
Drop in as a Claude-compatible endpoint — works with any tool that supports the Messages API.
export ANTHROPIC_API_KEY="mavis-xxxx..."
export ANTHROPIC_BASE_URL="https://mavis.io.vn"
curl $ANTHROPIC_BASE_URL/v1/messages \
-H "x-api-key: $ANTHROPIC_API_KEY" \
-H "anthropic-version: 2023-06-01" \
-H "content-type: application/json" \
-d '{
"model": "claude-opus-4.6",
"max_tokens": 1024,
"messages": [{ "role": "user", "content": "Hello" }]
}'