February 2026 Release
Release Type: Patch Release
Release Version: v0.2.0
GPU AI Service
New Features
Network Customization
You can now define custom TCP ports to access your GPUs outside the private network.
Bug Fixes
Enhanced SSH Connectivity
Resolved issues that intermittently prevented SSH access to GPU instances, ensuring reliable remote connections.
Improved GPU Selection
The GPU selection logic has been updated to accurately reflect the billing based on the exact VRAM available.
AI Inference
Model API
New Features
New LLM Models
Added support for more powerful LLM models, expanding the range of available options for inference.
Known Issues
Streaming Disabled: Streaming references are currently disabled for the AI Inference API.
Error Messages: When no GPUs are available in a selected region, the error message may incorrectly state "insufficient resources" instead of explicitly indicating unavailability.
Storage
New Features
Free Storage Tier
We have introduced a 100 GB free network storage tier to help you get started with your data and checkpoints without any additional costs.
Bug Fixes
Billing Inaccuracies Resolved
Fixed issues where disk costs were calculated incorrectly in the UI and ensured that attaching existing storage to a new instance does not result in double charging.
Platform Features
New Features
INR Payment Support
You can now make payments in INR (Indian Rupee) using UPI, Credit Cards and Netbanking options.
Integrated Support System
A new support and ticketing module has been integrated directly into the platform, making it easier to raise and track support requests.
Credits Visibility
The Credits and AI Inference overview pages are now accessible to all roles, providing better visibility into usage and balance.
Bug Fixes
Security & Login Enhancements
Resolved issues causing slow or failed OAuth logins. Additionally, access and refresh tokens are now issued only after both email and phone verification are complete.
Notes
Please report any issues via the new support ticketing system.
We have imporved our public documentation, you can check it here
Last updated