NVIDIA NIM
online
100+ curated AI models. $25 free credits.
LLMest. 1993 · Santa Clara, CA
95
/ 100 APIVault Score
// At a glance
Free Tier
$25 credits · 40 req/min · no card
Category
LLM
Credit Card
Not required
Last Verified
2m ago
// Free tier details
Available Models
Llama 3.3 70BMixtralNemotronStable Diffusion
Monthly Requests
$25 free credits (trial)
Rate Limit
40 requests/minute
No credit card needed
No phone verification
// Quick start
300">"text-purple-400">from openai 300">"text-purple-400">import OpenAI
client = OpenAI(
api_key=300">"YOUR_NVIDIA_KEY",
base_url=300">"https://integrate.api.nvidia.com/v1",
)
response = client.chat.completions.create(
model=300">"meta/llama-3.3-70b-instruct",
messages=[{300">"role": 300">"user", 300">"content": 300">"Hello."}],
)
print(response.choices[0].message.content)// Overview
NVIDIA's hosted API for running optimized NIM microservices — Llama, Mistral, Stable Diffusion, and more, all running on H100 GPUs. Generous $25 free starter credits.
// Pros
- $25 free credits is generous
- H100-powered speed
- 100+ models in one place
// Cons
- Credits expire
- Requires NVIDIA account
// Score breakdown
Reliability (35%) (from 2m ago health check)100/100
Free Tier Generosity (30%) (computed from quota, no-CC, no-phone fields)100/100
Documentation (20%) (human rating)86/100
Popularity (15%) (GitHub stars (log-normalised), or manual baseline)84/100
Methodology: apivault.dev/methodology
// Best for
Model evaluationH100-grade inferenceEnterprise AI
// Recent changes
May 12, 2026Added Llama 3.3 70B Nemotronadded