> ## Documentation Index > Fetch the complete documentation index at: https://docs.famulor.io/llms.txt > Use this file to discover all available pages before exploring further. # AI regions & models > Choose your workspace AI region, compare EU and US model endpoints, and understand SIP routing and infrastructure locations Choose where your workspace processes AI requests and see which models are available in each region. Famulor supports **Global**, **EU** and **US** regional settings. ## Choose your workspace AI region Open **Settings → Workspace → AI inference region**, choose **Global**, **EU** or **US**, then save. Workspace owners and admins can change this setting. | Setting | AI model availability | | - | - | | Global | Automatic routing without restricting AI requests to EU or US endpoints. | | EU | Models with an EU endpoint available through Famulor. | | US | Models with a US endpoint available through Famulor. | Workspace AI region selector with Global, EU and US options The region filters the models offered to your workspace. Your plan, assistant engine mode and language determine the remaining choices. Manual model selection requires **Fallbacks & Guardrails**; otherwise Famulor selects compatible models automatically. See [Models & voices](/assistants/models-and-voices). **Global** automatically selects the nearest available location for voice processing. ## SIP trunks in the EU and US Famulor SIP trunks are available in the **EU and US**. For assistant calls using phone numbers purchased through Famulor, saving the workspace region automatically switches inbound and outbound routing to the corresponding region: **EU → EU**, **US → US**. This automatic switch applies to Famulor-provided phone numbers used by assistants. [Your own SIP trunks](/telephony/sip-trunks), verified caller IDs and Loop phone routes keep their separate routing settings. ## Infrastructure locations | Component | Location | | - | - | | Database and authentication | Frankfurt, Germany (EU) | | Application hosting and server-side processing | Frankfurt, Germany (EU) | | Voice-agent workers | EU (Central Europe) and US (East) | | Famulor SIP trunks | EU and US, selected as described above | | AI model endpoints | EU, US or Global, according to the workspace region and model availability | Changing the workspace AI region does not move the database or application hosting out of Frankfurt. For named infrastructure providers, storage locations and data-processing details, see the [Trust Center](/support/trust-center). ## Supported models Catalog checked on **October 2, 2026**: **86 enabled models** — 28 language models, 21 speech recognition models, 28 speech generation models and 9 realtime speech models. **73** offer an EU endpoint and **85** offer a US endpoint through Famulor. **Yes** means the model is available with the workspace restricted to that region. **No** means it is not offered with that regional restriction; it may still be available with **Global**. The tables describe Famulor’s available catalog, not every model offered by each model family. **Gemini 3.5 Transcribe Live** is currently available only with **Global**. It has neither an EU nor a US regional endpoint in the Famulor catalog. ### Language models (LLM) | Model | EU endpoint | US endpoint | | - | - | - | | Gemma 4 31B | Yes | Yes | | Gemini 3.1 Flash Lite | Yes | Yes | | Gemini 3 Flash | No | Yes | | Gemini 3.5 Flash | Yes | Yes | | Gemini 3.1 Pro | No | Yes | | Gemini 3.5 Flash Lite | Yes | Yes | | Gemini 3.6 Flash | Yes | Yes | | Gemini 3.7 Flash | Yes | Yes | | Gemini 3.8 Flash | Yes | Yes | | GPT-5.6 Luna | Yes | Yes | | GPT-4.1 mini | Yes | Yes | | GPT-5.4 | Yes | Yes | | GPT-5.4 mini | Yes | Yes | | GPT-4.1 | Yes | Yes | | GPT-4.1 nano | Yes | Yes | | GPT-4o | Yes | Yes | | GPT-4o mini | Yes | Yes | | GPT-5 mini | Yes | Yes | | GPT-5 nano | Yes | Yes | | GPT-6 Luna | Yes | Yes | | GPT-5.4 nano | Yes | Yes | | GPT-5.5 | Yes | Yes | | ChatGPT Latest | No | Yes | | GPT-5 | Yes | Yes | | GPT-5.1 | Yes | Yes | | GPT-5.2 | Yes | Yes | | GPT-5.6 Sol | Yes | Yes | | GPT-5.6 Terra | Yes | Yes | ### Speech recognition (STT) | Model | EU endpoint | US endpoint | | - | - | - | | AssemblyAI Universal-3.5 Pro | Yes | Yes | | AssemblyAI Universal-3 Pro | Yes | Yes | | AssemblyAI Universal-Streaming | Yes | Yes | | Universal-Streaming Multilingual | Yes | Yes | | Universal-3.6 Pro Streaming | Yes | Yes | | Cartesia Ink Whisper | No | Yes | | Cartesia Ink 2 | Yes | Yes | | Deepgram Flux (EN) | Yes | Yes | | Deepgram Flux Multilingual | Yes | Yes | | Deepgram Nova-3 | Yes | Yes | | Deepgram Nova-2 | Yes | Yes | | Nova-2 Phone Call | Yes | Yes | | Nova-3 Medical | Yes | Yes | | Nova-3 Pharma | Yes | Yes | | Nova-2 Conversational AI | Yes | Yes | | Nova-2 Medical | Yes | Yes | | ElevenLabs Scribe v2 Realtime | Yes | Yes | | Gemini 3.5 Transcribe Live | No | No | | Soniox STT v5 (EU) | Yes | Yes | | xAI | Yes | Yes | | xAI Speech to Text 2 | Yes | Yes | ### Speech generation (TTS) | Model | EU endpoint | US endpoint | | - | - | - | | Cartesia Sonic-3.6 | Yes | Yes | | Cartesia Sonic-3.6-2026-08-27 | Yes | Yes | | Cartesia Sonic-3.5 | Yes | Yes | | Cartesia Sonic Latest | Yes | Yes | | Sonic 3 | Yes | Yes | | Sonic 3 (2026-01-12) | Yes | Yes | | Sonic 3 Latest | Yes | Yes | | Sonic 3.5 (2026-05-04) | Yes | Yes | | Sonic Preview | Yes | Yes | | Deepgram Aura-2 | Yes | Yes | | Deepgram Flux TTS | Yes | Yes | | Eleven Flash v2.5 | Yes | Yes | | Eleven Turbo v2.5 | Yes | Yes | | Eleven v4 Turbo | Yes | Yes | | Eleven Multilingual v2 | Yes | Yes | | Eleven v3 | Yes | Yes | | Eleven v4 | Yes | Yes | | Eleven v3 Conversational | Yes | Yes | | Fish Audio S2.1 Pro | No | Yes | | Fish Audio S2 Pro | No | Yes | | Inworld TTS-2 | No | Yes | | Inworld TTS 1.5 Max | No | Yes | | Inworld TTS 1.5 Mini | No | Yes | | Inworld Realtime TTS 2.0 Flash | No | Yes | | GPT-4o mini TTS | Yes | Yes | | Rime Coda | No | Yes | | Soniox TTS v2 | Yes | Yes | | xAI TTS | Yes | Yes | ### Realtime speech models | Model | EU endpoint | US endpoint | | - | - | - | | Gemini 3.8 Live | Yes | Yes | | Gemini 3.8 Live (extended thinking) | Yes | Yes | | Gemini 3.1 Flash Live | Yes | Yes | | Gemini 2.5 Flash Native Audio (December 2025) | Yes | Yes | | GPT-Realtime-2.1 | Yes | Yes | | GPT Realtime 2.1 Mini | Yes | Yes | | OpenAI Realtime Mini | Yes | Yes | | OpenAI Realtime 2 (reasoning) | Yes | Yes | | GPT Live | No | Yes | This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.