I am currently in the final stages of preparing for a large-scale media art exhibition opening this month. I need immediate clarification on the following points to finalize our operational workflow and ensure system stability.
1. ElevenLabs Voice Integration
I am currently using both “ElevenLabs Voices” and “ElevenLabs-Private Voices” (via my own API key). Does the usage of “Private Voices” count towards my Convai interaction limit?
Is there a way to track the consumption of Convai-provided ElevenLabs Voices separately from my Private Voices usage?
2. Flagship LLMs: Version Management
What is the definition of the “latest version” for Flagship LLMs on your platform?
If I use an older version, does it bypass the Flagship LLM interaction cap policy, or does it still count toward the limit regardless of the version?
3. Usage Tracking & Visibility
Is there a way to view real-time monthly interaction consumption with a breakdown by service/model?
Where can I find this detailed usage information in the dashboard?
As our exhibition launch is approaching rapidly, I would greatly appreciate your prompt assistance.
P.S. I have not received a reply, which is truly disappointing. This is a very important project for an upcoming exhibition, and I am in quite a hurry.
ElevenLabs Private Voices and usage
When you connect your own ElevenLabs API key and use ElevenLabs Private Voices, the TTS cost is taken from your ElevenLabs quota, not from Convai’s ElevenLabs allocation. You can track that usage directly in your ElevenLabs account.
Flagship LLM versions and limits
“Latest version” refers to the most recent model variants released by each provider. If you use non flagship models, their usage does not count toward the Flagship LLM interaction cap. Any Flagship model, regardless of minor version, will count toward the Flagship quota.
Usage tracking and breakdown
Right now, there is no per LLM / per service breakdown in the dashboard. You can see your overall usage and remaining quota on the Billing page under usage details, but it is not split by individual model.
Thank you for your support. I am preparing an interactive installation for an upcoming two-week exhibition and need clarification on the following points to ensure system stability.
1. Flagship Quota Definitions
The model selection menu lists gemini-2.5-flash as the latest, but also includes variations like gemini-2.5-flash-live(beta). Does gemini-2.5-flash-live(beta) count as a flagship model that consumes the Flagship quota?
Does using an older, non-flagship model effectively bypass the flagship interaction cap?
Could you provide an official, comprehensive list of which models currently count toward the “Flagship quota”?
2. Behavior When Quota is Exceeded
As there is no per-service usage breakdown in the dashboard, I am concerned about potential service disruptions. If the interaction limit is reached during the exhibition, will the characters immediately stop responding?
Is there an automated fallback system (such as a generic error message), or will the API simply return an error and silence the characters?
3. Mitigation Strategy
If usage approaches the limit, is it possible to request an immediate, temporary quota increase or plan upgrade to prevent service interruption?
Do you have any recommended best practices for maintaining long-term installations to avoid unexpected service stops?
P.S. I am in a real rush because I am preparing for an exhibition that starts tomorrow. As this is for a high-priority project, I would greatly appreciate your immediate assistance.
If your usage is getting close to the quota, you will automatically receive email notifications at different thresholds (for example, when you reach around 80%, 90%, 95%, and 100% of your quota). You can also monitor your remaining usage at any time from the Billing page in your dashboard.
If you see that your quota is about to run out, you can upgrade your plan from the Pricing page to avoid interruption. At the moment, there is no separate “quota top up” or extension for the same plan, but this is something we are working on. As a workaround, if you have another Convai account, you can also use that account’s API key in your Experience.
