Last updated
What is Simplismart used for in a Bolna voice agent?
Simplismart is an inference platform that optimises open models for speed, and it serves the LLM for a Bolna voice agent from either a shared endpoint or a dedicated deployment. Start on the shared endpoint to try a model on a call, then move to a dedicated endpoint for stable latency under production load or to run your own fine-tuned weights. Bolna connects to it as a custom LLM using the Simplismart preset.
How do I set up Simplismart with Bolna?
- Sign up at app.simplismart.ai and try an LLM in the Playground, for example Gemma 4 31B Instruct.
- Go to Settings → API Keys → Generate Key, name the key, choose a long expiry and copy it.
- For the shared endpoint, click Get API details in the Playground to copy the model ID, such as
google/gemma-4-31B-it. For a dedicated endpoint, deploy from the Model Marketplace and copy the URL and model name from the deployment's API tab. - In Bolna, open Settings → Integrations → LLM → Custom LLM → Add model, or choose + Add custom LLM from an agent's LLM Provider dropdown. Select Simplismart.
- Enter the model ID, set Endpoint URL to
https://api.simplismart.live(no/v1) or your dedicated URL, paste the key and set Family togemma, for example. - On the Intelligence tab choose custom and your Simplismart model, then save and test the agent.
The Simplismart setup guide in the Bolna docs explains when to move to a dedicated endpoint.
What is the difference between Simplismart shared and dedicated endpoints?
| Shared endpoint | Dedicated endpoint | |
|---|---|---|
| Infrastructure | Simplismart's shared pool | Isolated capacity reserved for you |
| Models | Pre-deployed models you can try in the Playground | Pre-optimised Marketplace models, or your own and fine-tuned models |
| Pricing | Per API call and usage | Deployment-based |
| Endpoint URL | https://api.simplismart.live, with no /v1 | Shown on the deployment's API tab |
Bolna sends only the Authorization header. If a dedicated deployment's snippet also needs an id header, ask Simplismart for an endpoint that doesn't.
What can you build with Simplismart and Bolna?
Agents on fine-tuned or private models. Only dedicated endpoints can run your own weights, so deploy there and register the endpoint as a custom model.
High-volume campaigns with steady latency. Move from shared to dedicated when time to first token gets uneven during busy hours or you need a latency SLA.
Quick open-model trials. Judge reply speed and quality in the Playground, then register the shared endpoint and place a web call.
Region-specific deployments. Ask Simplismart for a dedicated deployment near your callers. Nebius is another provider with both serverless and dedicated endpoints, and the custom LLM page lists every preset.