💳 Per-second billing
Lightning AI
AI Researchers, PyTorch Lightning Users, Collaborative Model Development
End-to-End MLOps✓ Free tier
⚙ Custom SDK
from$1.29
/ sec
Catalytic AI
🤖 Managed InferenceOn-premise and air-gapped LLM fine-tuning; private AI server builds
Catalytic AI builds and runs on-premise AI hardware for work that cannot go to a public cloud. We fine-tune models on corpora that must stay in the building, and build the machine the model runs on: three reference designs from a 16 GB entry node (PALLADIUM-16) to a 96 GB single-GPU workstation (PLATINUM-96) and a 192 GB dual-GPU training rig (RHODIUM-192). Every machine is specced against the model rather than a price list, sourced, built, burned in under sustained load, and documented for audit. Air-gapped handover is available: the machine can arrive having never been connected to the public internet. We also sell GPU hours on our own nodes (RTX PRO 6000 Blackwell 96 GB, NVIDIA DGX Spark 128 GB unified memory) for testing a workload before committing to a build. Based in New York; the build rack is in Troy, NY.
Supported Workloads
LLMImageEmbedding
Served Models
Qwen3-14B (fine-tuned)
Custom fine-tunes up to 100B+ (QLoRA)
SD3
FLUX.1
Data Privacy Policy
Data stays on-premise; not used for training; air-gapped option; client keeps weights
Custom API
Custom Integration. This provider requires their own specific SDKs or libraries to interact with the models. See official documentation.
Quick Start Snippet
Python
import requests
headers = {
'Authorization': 'Bearer YOUR_API_KEY',
'Content-Type': 'application/json'
}
data = {
'model': 'your-chosen-model',
'prompt': 'Hello, world!'
}
response = requests.post('https://catalyticgpu.com/v1/completions', headers=headers, json=data)| Website | Visit Official Site ↗ |
Billing Model
Usage-based
Usage-based pricing without fixed server costs.
Enterprise Singapore Allocates Over S$45 Million to Support AI Adoption in Advanced Manufacturing
Enterprise Singapore allocates S$45M to speed AI adoption in advanced manufacturing
EnterpriseSG to commit over S$45 million to help manufacturers adopt AI solutions
Singapore sets aside over S$45 million to help manufacturers adopt AI
Singapore allocates $35 million to accelerate AI adoption in advanced manufacturing
Community Discussions
0 CommentsJoin the Conversation
Sign in to ask questions, share insights, and connect with verified providers.
No discussions yet. Be the first to start the conversation!
Frequently Asked Questions
Catalytic AI uses a usage-based billing model. You pay only for what you use — no idle server costs.
Catalytic AI has its own API. Check their documentation for integration guides.
Catalytic AI supports LLM, Image, Embedding. Use the API to deploy custom models or use their pre-built endpoints.
Catalytic AI does not have a publicly listed free tier. Contact them for trial access or pilot pricing.
Explore More Options
By Region & GPU
More 🤖 Managed Inference Providers
💳 Per-token billing
Together AI
Finetuning Open Source Models, Serverless inference endpoints
LLMVisionEmbedding✓ Free tier
✓ OpenAI-compatible API
from$5.49
/ 1M tokens
💳 Per-token billing
Fireworks.ai
Uncompromising Speed and Precision Fireworks.ai was founded by former Meta…
LLMVision✓ Free tier
✓ OpenAI-compatible API
from$7.00
/ 1M tokens