- Joined
- Aug 19, 2025
- Messages
- 160
- Reaction score
- 64
- Points
- 28
Ally Just Got a Major Backend Upgrade — Powered by Nova Series AI
Hey everyone,
Quick but exciting update: we've swapped out Ally's backend for a newer, more capable model from our Nova AI series. This isn't just a behind-the-scenes tweak—it's a meaningful improvement to how Ally thinks, responds, and handles everything from casual chats to deep roleplay sessions.
What's live right now?
Why does this matter?
We built these models with real usage in mind. Whether you're using Ally for:
...you should notice smoother responses, better follow-through on complex prompts, and fewer "wait, what were we talking about?" moments.
We didn't rush this.
Before flipping the switch publicly, we ran Nova V4 Flash Thinking internally for a few months across a range of scenarios. The feedback was consistently positive: more reliable outputs, better handling of nuanced requests, and solid performance even during extended sessions. That gave us the confidence to roll it out now.
Try it yourself—no paywall.
What's coming next?
This is just the start. The full Nova platform launch is on the way at nova-ai.nabzclan.vip, and we're preparing:
We'd love to hear from you.
Try Ally with the new backend, poke around the API, and let us know what you think. What's working well? Where could it be smoother? Your input directly influences what we build next.
Thanks for sticking with us—and for helping shape what comes next.
P.S. If you run into any issues or have questions about the update, drop a note in this thread or reach out via the helpdesk. We're here.
Hey everyone,
Quick but exciting update: we've swapped out Ally's backend for a newer, more capable model from our Nova AI series. This isn't just a behind-the-scenes tweak—it's a meaningful improvement to how Ally thinks, responds, and handles everything from casual chats to deep roleplay sessions.
What's live right now?
- Nova V4 Flash Thinking is now active on Ally. It's faster, more coherent, and better at keeping track of context over longer conversations.
- Nova V4.1 Thinking is in the final stages—we'll share a release around June or July. Expect even sharper reasoning and more natural tone adaptation.
Why does this matter?
We built these models with real usage in mind. Whether you're using Ally for:
- Everyday help—writing, research, brainstorming, or just chatting
- Roleplay and creative storytelling—where consistency and character voice matter
- Testing integrations via our API—clean endpoints, clear docs, no fluff
...you should notice smoother responses, better follow-through on complex prompts, and fewer "wait, what were we talking about?" moments.
We didn't rush this.
Before flipping the switch publicly, we ran Nova V4 Flash Thinking internally for a few months across a range of scenarios. The feedback was consistently positive: more reliable outputs, better handling of nuanced requests, and solid performance even during extended sessions. That gave us the confidence to roll it out now.
Try it yourself—no paywall.
- Chat with Ally directly: helpdeskai.nabzclan.vip/chat
- Developer? Hook into the API: API Docs & Quick Start
When calling the endpoint, use this model name: NovaV4_flash_Thinking
What's coming next?
This is just the start. The full Nova platform launch is on the way at nova-ai.nabzclan.vip, and we're preparing:
- Additional model variants tuned for specific use cases
- More developer tools and customization options
- Community-driven features based on your feedback
We'd love to hear from you.
Try Ally with the new backend, poke around the API, and let us know what you think. What's working well? Where could it be smoother? Your input directly influences what we build next.
Thanks for sticking with us—and for helping shape what comes next.
— The NabzClan Team
P.S. If you run into any issues or have questions about the update, drop a note in this thread or reach out via the helpdesk. We're here.
