مرّر فوق النص للترجمة · Hover text to translate · Tap on mobile
NVIDIA NIM
Free OpenAI-compatible inference across 100+ models on DGX Cloudاستدلال مجاني متوافق مع OpenAI لأكثر من 100 نموذج على DGX CloudFree OpenAI-compatible inference across 100+ models on DGX Cloud
NVIDIA’s free inference platform at build.nvidia.com. Join the NVIDIA Developer Program to get API access to 100+ leading models (Nemotron, GLM, MiniMax, DeepSeek and more) served on DGX Cloud. Every model exposes an OpenAI-compatible endpoint, so switching models is just a string change. New developers get 1,000-5,000 free inference credits with roughly 40 requests/minute per model - perfect for prototyping and building agents.منصة الاستدلال المجانية من NVIDIA على build.nvidia.com. انضم لبرنامج مطوري NVIDIA واحصل على وصول لـ 100+ نموذج (Nemotron، GLM، MiniMax، DeepSeek) على DGX Cloud مع نقطة متوافقة مع OpenAI و 1000-5000 رصيد مجاني.NVIDIA’s free inference platform at build.nvidia.com. Join the NVIDIA Developer Program to get API access to 100+ leading models (Nemotron, GLM, MiniMax, DeepSeek and more) served on DGX Cloud. Every model exposes an OpenAI-compatible endpoint, so switching models is just a string change. New developers get 1,000-5,000 free inference credits with roughly 40 requests/minute per model - perfect for prototyping and building agents.
Open the NVIDIA model catalogافتح كتالوج نماذج NVIDIAOpen the NVIDIA model catalog
Go to build.nvidia.com/models to browse the full catalog of supported models, filters, and playgrounds.اذهب إلى build.nvidia.com/models لاستعراض كل النماذج.Go to build.nvidia.com/models to browse the full catalog of supported models, filters, and playgrounds.
2
Join the NVIDIA Developer program (free)انضم لبرنامج المطورين (مجاني)Join the NVIDIA Developer program (free)
Create an account on build.nvidia.com and join the NVIDIA Developer Program. This unlocks the hosted NIM APIs and can award 1,000-5,000 free inference credits.أنشئ حساباً وانضم لبرنامج NVIDIA Developer للحصول على 1000-5000 رصيد مجاني.Create an account on build.nvidia.com and join the NVIDIA Developer Program. This unlocks the hosted NIM APIs and can award 1,000-5,000 free inference credits.
3
Generate your NVIDIA API keyأنشئ مفتاح APIGenerate your NVIDIA API key
In your account, generate an API key that starts with nvapi-. Keep it in an env var for your app or agent.أنشئ مفتاحاً يبدأ بـ nvapi- واحتفظ به في متغير بيئة.In your account, generate an API key that starts with nvapi-. Keep it in an env var for your app or agent.
4
Call the OpenAI-compatible endpointاستدعِ النقطة المتوافقةCall the OpenAI-compatible endpoint
Use https://integrate.api.nvidia.com/v1 as your base URL and pass the model ID of any catalog model. Your existing OpenAI SDK code works with just a base URL + key swap.استخدم https://integrate.api.nvidia.com/v1 كنقطة أساس ومرر معرف النموذج.Use https://integrate.api.nvidia.com/v1 as your base URL and pass the model ID of any catalog model. Your existing OpenAI SDK code works with just a base URL + key swap.
5
Prototype within the free tierجرّب ضمن الحد المجانيPrototype within the free tier
Each model is rate-limited (roughly 40 req/min) and credits are consumed per request. Perfect for prototypes, agents, and experiments - move to paid or self-hosted NIM for production.كل نموذج محدود بنحو 40 طلب/دقيقة - مثالي للنماذج الأولية والوكلاء.Each model is rate-limited (roughly 40 req/min) and credits are consumed per request. Perfect for prototypes, agents, and experiments - move to paid or self-hosted NIM for production.
