DrAI aggregates 40+ models behind one OpenAI-compatible endpoint: GPT-5 series (gpt-5.6-sol, gpt-5.4, gpt-5.4-mini, gpt-5.2), Claude 4 (Opus/Sonnet), DeepSeek R1/V3, Gemini 2.5 Pro, Qwen, GLM, Kimi, Midjourney image models, and Embedding vectors. New models are added continuously. Check the live model list in your dashboard for current IDs and per-model pricing.

Sign in at ai.dr-ai.top, open the API Keys section in your dashboard, and click Create Key. Your key is displayed only once, so copy and store it securely. Each account can create multiple keys for different projects or environments, and you can revoke any key instantly from the same page.

Three plans: Free (basic models, daily quota, no credit card required), Pro at $9.99/month, and Max at $29.99/month, with yearly billing at 12x the monthly price. Pay-as-you-go credit top-ups are also available for variable workloads. API calls are billed per token at the transparent per-model rates shown on the pricing page.

All requests are encrypted in transit with HTTPS/TLS, and API keys are isolated per account, never shared between tenants. Usage logs and stats are visible only in your own dashboard. We do not train models on your prompts or outputs. For production, call the API server-side and never embed keys in client code.

Limits scale with your plan: Free has the lowest per-minute caps, Pro a middle tier, and Max the highest. Exceeding a limit returns HTTP 429 with a Retry-After header, so implement exponential backoff in your client. Upgrading raises limits immediately; limits reset on a rolling per-minute window. Your current caps and live usage are shown on the dashboard usage page.

Change only two values: set base_url to https://ai.dr-ai.top/v1 and swap the API key for a DrAI key. All endpoints follow the OpenAI Chat Completions format, so streaming, function calling, and JSON mode keep working unchanged. Most codebases migrate in under 10 minutes with zero logic changes. The /v1/models endpoint lists every model available to your key, so you can verify access right after switching.

Yes. DrAI delivers full SSE streaming identical to OpenAI's stream:true behavior, with token-level deltas in real time. First token typically arrives within 1-2 seconds for most models. Streaming also avoids client-side timeouts on long generations and greatly improves perceived latency in chat UIs, making DrAI a drop-in replacement for latency-sensitive production chat and agent workloads.

Any OpenAI-compatible SDK works: the official Python openai package, Node.js/TypeScript, Go, Java, Ruby, and PHP clients, plus frameworks such as LangChain, LlamaIndex, Dify, and Cursor. Point the SDK's base_url at https://ai.dr-ai.top/v1 with your DrAI key. No special client is required. Both sync and async clients are supported, and tool-calling examples in every language are in the API docs.

Yes. One key calls every model in the catalog. Build fallback chains such as gpt-5.4 to claude-4-sonnet to deepseek-r1, so a busy or degraded model rolls over automatically to the next. Per-model routing, retries, and usage breakdowns are all visible in the dashboard, making it easy to build cost-aware routing: cheap models for bulk work, frontier models for hard reasoning.

Yes. The Free plan gives immediate access to basic models with a daily quota and requires no credit card, enough for evaluation, small scripts, and hobby projects. Upgrade to Pro ($9.99/mo) or Max ($29.99/mo) for higher limits, premium models, and priority throughput. There are no hidden fees or automatic upgrades, so you stay on Free until you choose to switch.