At Inferless, we're committed to providing our users with the most efficient and powerful tools for serverless GPU inference. This month, we're excited to announce the launch of our new UI, designed to streamline workflows and boost productivity for our customers.
Our latest update addresses key challenges faced by our users, including focusing on simplifying model deployment, resource management, and cost optimization. By reimagining our interface and underlying architecture, we've created a more intuitive, responsive, and powerful platform to improve your experience.
Key improvements
-
Simplified deployment process
Deploy custom models in one page, reducing steps from 5 to 1.
-
Enhanced visibility
Clear visuals for GPU specs, pricing, and performance metrics.
-
Intuitive testing
New playground feature for quick model testing.
-
Improved debugging
Separated inference and build logs for easier troubleshooting.
-
AI-powered support
Introducing Ask Docs AI for instant guidance.
Streamlined deployment process
Our new UI simplifies model deployment into one page:
- Select your model source (Hugging Face, GitHub, etc.)
- Choose your GPU specifications
- Configure deployment settings
- Deploy with a single click
- Connect your provider account to access your models directly from the dropdown menu.
Enhanced visibility and control
Clear GPU specifications
We now provide detailed information for each GPU type, including:
- GPU RAM
- vCPUs
- RAM
- Cost per second
Transparent pricing
Get a clear monthly price estimation as you select your GPU, helping you make cost effective decisions.
Comprehensive performance metrics
Monitor your model's performance with detailed graphics showing:
- Total calls
- Cold-start times
- GPU/CPU utilization
Playground: test your models instantly
Our new playground feature allows you to test your models directly in the console, streamlining your development process.
I/O tracking
View a summary of your last 20 calls, providing valuable insights into your model's inputs and outputs.
Improved logging
We've separated inference logs from build logs, making debugging faster and more efficient.
AI-powered support
Introducing Ask Docs AI, our new AI-powered chatbot. Get instant support and guidance, trained on our documentation and common deployment scenarios.
Getting started
We're excited to see how these new features enhance your serverless GPU inference workflows. As always, we welcome your feedback as we continue to improve our platform.
We’re hosting a live Townhall on Aug 28th to show you all the updates and get answers from our founders as they demo the new deployment process and guide you on how to save your time for future deployments – click here to RSVP.