Model Database is a unified, OpenAI-compatible API for hundreds of large language models. Access Claude, GPT, Gemini, Llama, Mistral and DeepSeek through a single endpoint — no infrastructure, no per-provider contracts. Sign up and ship in minutes.
Free account · API key in seconds · Pay only for what you use
# Already use the OpenAI SDK? Change one line: the base URL. curl https://modeldatabase.com/v1/chat/completions \ -H "Authorization: Bearer $MDB_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "anthropic/claude-opus-4.8", "messages": [{"role":"user","content":"Hello!"}], "stream": true }'
Call any of these by passing its ID in the model field — or browse the full
catalog of hundreds more. One key, one bill, every provider.
Anthropic's most capable model — deep reasoning, agentic workflows, and frontier coding.
Near-flagship quality at a fraction of the cost and latency. A great everyday default.
Versatile, multimodal, and widely supported — a dependable choice for general chat.
Fast and inexpensive for high-volume classification, extraction, and routing.
Low-latency multimodal model with a very large context window for long documents.
Capable open-weight workhorse with excellent price-to-performance.
Strong reasoning and multilingual performance from the European frontier lab.
Exceptional value for coding and chat, with surprisingly strong reasoning.
Every gateway says “hundreds of models, one key.” So do we — but that's table stakes. What actually sets us apart: every price is published, your spend is capped, and we never store your prompts.
Every model's per-token rate is published and sortable — the price you see is the price you pay, margin included. No opaque markups.
Credit is prepaid, so you can never overspend — and a per-request cost cap stops a runaway prompt cold. Predictable by design.
We don't store the content of your prompts or the models' responses, and we never train on them. Only usage metadata for your bill. See our privacy policy.
Fully OpenAI-compatible /v1/chat/completions with streaming, tool calling, and JSON mode. Keep your SDK — just point the base URL at us.
Switch between Claude, GPT-class, Gemini, Llama, Mistral, DeepSeek and hundreds more by changing a single model string.
Top up with a card via Stripe, or pay in BTC, ETH, or USDT. No subscriptions, no seats, no minimums — credit never expires.
Sign up with an email and password. You'll get an mdb_live_ API key instantly.
Top up your balance from the dashboard. Credit works across every model — no per-provider setup.
Set base URL to modeldatabase.com/v1, pass your key as a Bearer token, and send requests.
Prepay a credit pack, then pay only for the tokens you actually use. Credit never expires.
Perfect for prototyping and testing.
For apps moving into production.
For high-volume workloads.
All packs draw down per request at published, per-model token rates. See every model’s price →
Yes. The endpoint mirrors /v1/chat/completions, including streaming. In most codebases you only change the base URL to https://modeldatabase.com/v1 and use your Model Database API key.
Hundreds — spanning Claude, GPT-class models, Gemini, Llama, Mistral, DeepSeek, Qwen and more. You pick a model by passing its identifier in the model field. Call GET /v1/models with your key to list everything available.
You prepay a credit pack ($10, $50, or $200) from your dashboard — with a card via Stripe, or in crypto (BTC, ETH, or USDT). Each request draws down your balance based on the tokens used, and every response includes headers showing the cost of the call and your remaining balance.
Seconds. Create an account, copy your mdb_live_ key, add credit, and make your first call. No sales calls, no waitlist.
Three things most aggregators won't give you straight: we publish every model's price (sortable, margin included — no guessing your markup), your spend is capped by prepaid credit plus a per-request safety limit, and we never store the content of your prompts or responses. Same one-line migration, fewer unknowns.
Yes. Create as many keys as you like, label them, and revoke any of them instantly from your dashboard if one is compromised or no longer needed.
Join developers shipping faster with one API for every model.
Get your API key →