Changelog
Track the latest updates, new features, and improvements in Novisurf.
Recent Updates
Stay informed about Novisurf's latest releases. You receive new features, performance improvements, and bug fixes to enhance your AI inference experience. Check the tags for quick categorization: feature for new capabilities, bugfix for fixes, breaking for changes requiring updates, and improvement for optimizations.
Upgrade to the latest version for the best performance and security. Run npm update novisurf-sdk or check your dashboard at https://platform.novisurf.top.
New Features
- Added support for Unseen, our first proprietary AI model optimized for RAG workflows.
- Introduced automatic context retrieval from uploaded knowledge bases during inference.
Improvements
- Reduced average response times to
<500msusing Cerebras and Groq backends. - Enhanced rate limits for production workloads, now supporting
>10k RPMfor paid tiers.
Bug Fixes
- Fixed intermittent timeouts in high-concurrency scenarios.
- Resolved compatibility issues with Vercel AI SDK v3.2+.
New Features
- Pay-as-you-go billing with per-token pricing and no markups.
- Dashboard integration for monitoring usage and managing API keys.
Bug Fixes
- Corrected token counting inaccuracies for certain models.
- Patched potential race conditions in concurrent RAG queries.
Improvements
- Better error messages for invalid
baseURLconfigurations.
Initial Release
- Launched unified inference gateway compatible with OpenAI API spec.
- Built-in RAG support: Upload documents via dashboard for automatic retrieval.
Breaking Changes
baseURLchanged tohttps://api2.novisurf.top/v1(update your SDK config).
Key Features
- Seamless integration with Vercel AI SDK, LangChain, and more.
- High-capacity rate limits from day one.
Upgrade Guide
Follow these steps to upgrade your Novisurf integration safely.
Check Current Version
Verify your SDK version:
npm list novisurf-sdk
Update SDK
npm install novisurf-sdk@latest
yarn add novisurf-sdk@latest
Update Config
Ensure your baseURL points to https://api2.novisurf.top/v1:
import { createOpenAI } from '@ai-sdk/openai';
const novisurf = createOpenAI({
apiKey: `YOUR_API_KEY`,
baseURL: `https://api2.novisurf.top/v1`,
});
from openai import OpenAI
client = OpenAI(
api_key="YOUR_API_KEY",
base_url="https://api2.novisurf.top/v1"
)
Test Integration
Run a simple inference test and monitor your dashboard.
For breaking changes, review the migration guide in your dashboard. Join our Discord for support: https://go.novisurf.top/join.
Explore Further
Last updated today
Built with Documentation.AI