# Lava > Lava is an AI billing and gateway platform. One API for routing, metering, pricing, and spend controls across 29+ AI providers and 530+ models. Ship AI products and get paid without building billing infrastructure. - Website: https://www.lava.so - Documentation: https://www.lava.so/docs - Pricing: https://www.lava.so/pricing - Contact: https://www.lava.so/contact - GitHub: https://github.com/lavapayments ## Products ### Lava Gateway URL: https://www.lava.so/products/gateway Lava Gateway is a unified API proxy for AI providers. You replace your AI provider's base URL with the Lava Gateway endpoint, and your existing code and SDKs work without changes. Lava routes requests to the correct provider, meters every token in real time, and gives you a unified dashboard across all providers. **Key features:** - Access 530+ models across 29+ AI providers through a single API endpoint - OpenAI-compatible API format works as a drop-in replacement for existing SDKs - Real-time metering with token-level granularity on every request - Sub-5ms latency overhead - Automatic failover when a provider is unavailable - Cost tracking broken down by model, provider, feature, and end user - Zero lock-in, switch providers without changing application code **Supported providers include:** OpenAI, Anthropic, Google Gemini, Mistral, xAI (Grok), DeepSeek, Groq, Together AI, Cohere, Perplexity, Fireworks AI, Cerebras, SambaNova, Nebius, Moonshot AI, ElevenLabs, Amazon Bedrock, and more. Also supports non-AI API services like Serper (web search, news, places, scholar). **How it works:** 1. Replace your AI provider base URL with `https://api.lava.so/v1/forward/{provider}/` 2. Use your existing OpenAI SDK or HTTP client 3. Every request is automatically metered with token counts, costs, and latency 4. View unified analytics across all providers in the Lava dashboard ### Lava Monetize URL: https://www.lava.so/products/monetize Lava Monetize is a billing platform for AI products. It handles payment collection, wallet-based billing, real-time balance tracking, and usage dashboards so you can launch any pricing model without building billing infrastructure. **Key features:** - Hosted checkout for your AI product (embed with a few lines of code) - Prepaid wallet system where users load funds and spend them on AI usage - Real-time balance tracking and automatic deductions per request - Configurable pricing with per-model markup control - Spending limits and usage alerts for end users - Customer-facing usage dashboards - Subscription billing with usage allowances - Automatic top-ups when wallet balance is low **Supported pricing models:** - Prepaid credits (users buy balance, spend on usage) - Subscriptions (monthly/annual plans with usage limits) - Usage-based billing (pay-as-you-go metered pricing) - Hybrid models (subscription + overage charges) **How it works:** 1. Create a product and pricing configuration in the Lava dashboard 2. Embed the hosted checkout in your application 3. Users fund their wallet through the checkout 4. Route AI requests through Lava Gateway, which automatically deducts costs from user wallets 5. Users see real-time balance, usage breakdown, and spending history ### Lava AI Spend URL: https://www.lava.so/products/ai-spend Lava AI Spend gives teams budget-controlled API keys (spend keys) that work with any OpenAI-compatible AI tool. **Key features:** - Generate spend keys that work with Cursor, Raycast AI, Windsurf, Continue.dev, Cline, Aider, and any OpenAI-compatible tool - Per-key budgets with daily, weekly, monthly, or lifetime spending limits - Model restrictions per key (control which AI models each key can access) - Real-time cost tracking across all keys - Automatic cutoff when spending limits are reached - One key accesses all 530+ models across 29+ providers **How it works:** 1. Create a Lava account and fund your wallet 2. Generate a spend key with your desired budget and model restrictions 3. Paste the key into any tool that accepts an OpenAI or Anthropic API key 4. Set the API base URL to `https://api.lava.so` 5. Track spending in real time from the Lava dashboard ### Service Catalog URL: https://www.lava.so/services API: https://api.lava.so/v1/services The Lava Service Catalog is a public, searchable directory of every model and API service available through the gateway. It includes pricing, capabilities, provider details, and whether each entry is managed by Lava or requires bring-your-own-key (BYOK). **Key features:** - Browse 530+ models and API services across 29+ providers - Filter by type (AI models vs API services), access model (managed vs BYOK), or search by name - View per-model pricing (per 1M tokens, per request, per minute, etc.) - See capabilities per model (chat, streaming, embeddings, image generation, text-to-speech, web search) - Public API endpoint at `GET /v1/services` returns the full catalog as JSON, no authentication required **Providers currently in the catalog:** OpenAI (72 models), Together AI (68), Mistral (47), Amazon Bedrock (44), Novita AI (36), xAI (29), Parasail (27), DeepInfra (23), Nebius (22), Anthropic (15), Inference.net (15), Google (14), Groq (12), Moonshot AI (12), ElevenLabs (11), kluster.ai (10), SambaNova (10), Cohere (10), Chutes (10), Hyperbolic (9), Fireworks (9), Cerebras (6), Serper (6 API services), Perplexity (4), Vercel (3), DeepSeek (2), GMI Cloud (2), Pearch (2, BYOK), Retell (1). ## Pricing URL: https://www.lava.so/pricing Lava charges a 1.9% service charge on transactions processed through the platform. Your first $10,000 in processed transactions is free. There are no monthly fees, per-seat charges, or minimum commitments. AI provider costs are passed through at cost with no markup. ## Frequently Asked Questions ### What is Lava? Lava is an AI billing and gateway platform. It provides two core products: Lava Gateway (a unified API proxy for 29+ AI providers) and Lava Monetize (a billing platform for AI products). Together, they let you route AI requests through a single endpoint and monetize your AI product without building billing infrastructure. ### How does the AI gateway work? You replace your AI provider's base URL with the Lava Gateway endpoint. Your existing OpenAI SDK or HTTP client works without code changes. Lava routes requests to the correct provider, meters every token in real time, and provides unified analytics across all providers. ### What AI providers does Lava support? Lava supports 29+ providers and 530+ models, including OpenAI, Anthropic (Claude Opus, Sonnet, Haiku), Google (Gemini 2.5 Pro, Flash), Mistral, xAI (Grok), DeepSeek, Groq, Together AI, Cohere, Amazon Bedrock, and more. Lava also supports non-AI API services like Serper for web search. See the full catalog at https://www.lava.so/services. ### How do I add billing to my AI product? With Lava Monetize, you add a few lines of code to embed a hosted checkout. Users fund a prepaid wallet and their balance decreases in real time as they use AI features. Lava handles payment collection, balance management, spending limits, and usage dashboards. ### What pricing models can I use? Lava Monetize supports prepaid credits, subscriptions, usage-based billing, and hybrid models. You configure your pricing in the dashboard and Lava handles the rest. ### What is a spend key? A spend key is an API key from Lava that works with any OpenAI-compatible tool and carries its own spending limits, model restrictions, and usage tracking. It routes requests through Lava Gateway, so you get cost tracking and budget enforcement automatically. ### How much does Lava cost? 1.9% service charge on processed transactions. First $10,000 free. No monthly fees, no per-seat charges. AI provider costs are passed through at cost. ### What is the latency overhead? Lava Gateway adds less than 5 milliseconds of latency to requests. Since AI model responses typically take hundreds of milliseconds to seconds, the gateway overhead is negligible. ## Technical Documentation - [Gateway Quickstart](https://www.lava.so/docs/quickstart): Get started with the AI gateway in minutes - [API Reference](https://www.lava.so/docs/api-reference): REST API endpoints for subscriptions, usage, and billing - [Node.js SDK](https://www.lava.so/docs/sdks/nodejs): Official Node.js SDK (`@lavapayments/nodejs`) - [Checkout SDK](https://www.lava.so/docs/sdks/checkout): Frontend SDK for embedded checkout (`@lavapayments/checkout`) - [AI Spend Keys Guide](https://www.lava.so/docs/guides/ai-spend-keys): Set budget-controlled API keys - [Spend Key Setup Guides](https://www.lava.so/docs/guides/spend-key-setup-guides): Connect spend keys to Claude Code, Cursor, Raycast, Cline, and more - [Checkout Integration](https://www.lava.so/docs/guides/checkout): Add hosted checkout to your application - [Service Catalog](https://www.lava.so/services): Browse all 530+ models and API services with pricing, capabilities, and provider details - [Service Catalog API](https://www.lava.so/docs/api-reference/list-services): Public API endpoint returning every available model and service with pricing ## Blog: The Molten Stack URL: https://www.lava.so/blog Lava's blog covers AI pricing, billing, gateways, and infrastructure. ### How to Set Up Billing for Your AI App: A Step-by-Step Guide URL: https://www.lava.so/blog/how-to-set-up-ai-billing A practical 5-step guide to adding billing to your AI product. Covers pricing models, metering, payment collection, spending controls, and usage dashboards. ### What Is an AI Gateway? A Guide for Developers URL: https://www.lava.so/blog/what-is-ai-gateway An AI gateway is a proxy between your app and AI providers. Learn why you need one, how they work, and what to look for when choosing a gateway. ### The Best AI Tools to Use with Lava Spend Keys URL: https://www.lava.so/blog/best-ai-tools-spend-keys A practical guide to the best AI coding assistants, chat interfaces, agent frameworks, and automation tools you can use with a single Lava spend key. ### The 5 Best AI Gateways URL: https://www.lava.so/blog/best-ai-gateways Compare Lava, Portkey, LiteLLM, Helicone, and OpenRouter. A practical breakdown of features, pricing, and which AI gateway fits your stack. ### The 5 Best AI Billing Platforms URL: https://www.lava.so/blog/best-ai-billing-platforms Compare Lava, Stripe, Orb, Metronome, Autumn, and Flowglad for AI billing. Real-time enforcement, pricing, and which platform fits your needs. ### The 5 Best AI Cost Tracking Tools URL: https://www.lava.so/blog/best-ai-cost-tracking Compare Lava, Helicone, LangSmith, Portkey, and Keywords AI for tracking LLM costs. Features, pricing, and which tool fits your workflow. ### The 5 Best AI Spend Management Tools URL: https://www.lava.so/blog/best-ai-spend-management Compare Lava, Helicone, Vantage, Portkey, and CloudZero for managing AI spend. Budget controls, alerts, and which tool prevents cost overruns. ### Pricing AI Agents: 5 Models for Charging for Autonomous AI URL: https://www.lava.so/blog/pricing-ai-agents AI agents are unpredictable by nature. Compare per-task, per-action, outcome-based, time-based, and credit-based pricing to find the right fit. ### How to Bill Your Users for AI: A Guide for Platform Builders URL: https://www.lava.so/blog/billing-end-users-ai You pay per-token to AI providers and need to pass those costs to your users. Here is what it takes to build platform billing for AI. ### Outcome-Based Pricing for AI: When It Works and When It Doesn't URL: https://www.lava.so/blog/outcome-based-pricing-ai Outcome-based pricing sounds perfect in theory. Here is an honest look at when it works, when it falls apart, and what to do instead. ### Credit-Based Pricing for AI Products: A Practical Guide URL: https://www.lava.so/blog/credit-based-pricing-ai Prepaid credits are becoming the default pricing model for AI products. How they work, who uses them, and how to implement them right. ### Usage-Based Billing for AI: What You Actually Need to Build URL: https://www.lava.so/blog/usage-based-billing-ai A practical guide to the four systems behind usage-based billing for AI products: metering, rating, collection, and visibility. ### The Hidden Costs of Running AI in Production URL: https://www.lava.so/blog/hidden-costs-ai-production LLM API calls are just the tip of the iceberg. A practical breakdown of the real costs behind AI products, from token waste to billing infrastructure. ### AI Pricing Models: Which One Fits Your Startup? URL: https://www.lava.so/blog/ai-pricing-models Compare subscription, usage-based, outcome-based, and hybrid pricing models to find the right fit for your AI product. ### The Startup Founder's Guide to Smarter Pricing URL: https://www.lava.so/blog/founders-guide-pricing Practical pricing principles for AI and SaaS founders, from value-based pricing to tier design and iterative experimentation. ### Choosing the Right LLM: A Guide to GPT, Claude, Gemini, and More URL: https://www.lava.so/blog/choosing-the-right-llm A practical breakdown of GPT, Claude, Gemini, Llama, and Mistral to help you pick the best large language model for your use case.