Skip to content
OpenRouter

OpenRouter LLM gateway for Bolna AI voice agents

OpenRouter lets a Bolna voice agent use models from OpenAI, Anthropic, Google, Mistral, Meta and DeepSeek through one API key, by setting the agent's LLM to an OpenRouter model slug such as openai/gpt-4o-mini. Pick it when you want to compare several models, keep a fallback ready, or avoid managing a separate account with each provider.

How OpenRouter fits in the stack.

  • Best for

    Reasoning, policy handling, tool use and response generation.

  • Use this layer when

    You want to tune quality, cost, latency or function-calling behaviour.

  • Connects to

    Transcribed caller input upstream, and the voice or actions downstream.

The LLM is the
reasoning layer.

OpenRouter decides how the agent reads intent, picks the next action, calls your tools and writes the reply. It's the decision engine inside the call.

  1. Telephony

    Phone network

    • Exotel
    • Plivo
    • Twilio
  2. STT

    Listener

    • Deepgram
    • Sarvam AI (STT)
    • Soniox
  3. LLM · You are here

    Reasoning

    • OpenRouter
  4. TTS

    Voice

    • ElevenLabs
    • Sarvam AI (TTS)
    • Cartesia
  5. Tools

    Actions

    • Cal.com
    • n8n
    • AiSensy

How do I set up OpenRouter with Bolna?

Which OpenRouter models does Bolna recommend for voice?

Model slugContextBest for
openai/gpt-4o-mini128KBest OpenAI balance through OpenRouter
openai/gpt-4o128KHigher-quality OpenAI
anthropic/claude-haiku-4.5200KFast, cost-efficient Claude
anthropic/claude-sonnet-4.1200KHigher-quality Claude
google/gemini-3-flash1MFast Gemini with good multilingual support
google/gemini-3-flash-lite1MCheapest large-context option
mistralai/mistral-small32KFast, low-cost European option
meta-llama/llama-3.1-70b128KOpen-weight, cost-effective
deepseek/deepseek-chat1MVery low cost

The OpenRouter setup guide in the Bolna docs lists every setting and a JSON config you can paste into the API.

The full catalogue is at openrouter.ai/models.

What can you build with OpenRouter and Bolna?

  • Model comparisons

    Run the same agent on different slugs to see which model gives the best mix of cost, speed and quality for your calls.

  • Agents with a fallback model

    Configure fallback routing in OpenRouter so a support line has a second model ready if the primary provider is unavailable.

  • Multilingual agents

    Use google/gemini-3-flash for agents that need to speak Hindi or other Indian languages.

  • Teams without provider accounts

    Reach Claude, Gemini, Llama and Mistral through one key instead of signing up with each vendor. If you later need a direct connection, see the Anthropic integration.

Other providers in this layer.

See all 18 LLMs
  • Perplexity can be the LLM behind a Bolna voice agent. Set the provider to perplexity in the agent config and Bolna routes calls to your chosen Perplexity model.

  • Simplismart serves speed-optimised open models to a Bolna voice agent on a shared endpoint or a dedicated deployment, including your own fine-tuned weights.

  • Together AI gives a Bolna voice agent a large serverless catalog of open models such as Llama, Qwen, DeepSeek and MiniMax, billed per token with no provisioning.

Keep exploring the voice stack.

See where OpenRouter fits in your workflow.

Walk through provider choice, stack trade-offs and the exact workflow you want Bolna to automate.

Ready for enterprise.

Secure, compliant and dependable, from the first pilot to full scale.

Contact Sales
  • 99.9% uptime

    Built for your busiest hour. Even if the model fails, the call doesn’t

  • Support and SLAs

    A dedicated team, with service levels shaped around your business

  • Security & compliance

    Data residency, encryption, access controls, full audit trails and more

  • Deploy in your cloud

    Run Bolna in your own cloud or on-premise, your data remains yours

Certified by

  • ISO 27001
  • DPDP compliant

Frequently Asked Questions

Very little. The Bolna docs put routing overhead at a few milliseconds, which is negligible for most agents. If your agent is very latency-sensitive, benchmark it against a direct provider connection.