Skip to content

Gateways

Fontana provides built-in support for a wide range of LLM gateways, from unified routers such as OpenRouter and LLM API to direct vendor APIs and cloud-hosted foundation models on AWS Bedrock. All calls route through one gateway abstraction with Bring Your Own Key (BYOK): you supply the API keys your organisation holds, and administrators approve which gateways and models are reachable in production. Embeddings run on a single active profile you select in Admin → Models → Embeddings: the in-cluster TEI service by default (air-gapped, no key required), or one of the external gateways using your BYOK key.

For retention and no-training posture on each gateway, see Zero Data Retention (ZDR).

IconGatewayExamplesDocumentation
OpenRouter300+ models via one keyopenrouter.ai/docs
LLM APISecondary unified gatewayllmapi.ai/docs
Merge GatewayUnified routing and failoverdocs.merge.dev/merge-gateway
RequestyAI gateway and LLM routerdocs.requesty.ai
OpenAIGPT / o-seriesplatform.openai.com/docs
AnthropicClaudedocs.anthropic.com
GoogleGeminiai.google.dev
GroqFast LPU inferenceconsole.groq.com/docs
MistralMistral Large / Small / Pixtraldocs.mistral.ai
NVIDIANemotron and hosted LLMs (NIM)build.nvidia.com/docs
Together AI200+ open modelsdocs.together.ai
DeepInfraOpen-source LLMsdeepinfra.com/docs
Fireworks AIServerless open modelsdocs.fireworks.ai
BasetenFrontier open modelsdocs.baseten.co
CerebrasUltra-fast inference (Llama, Qwen, GPT-OSS, GLM)inference-docs.cerebras.ai
xAI (Grok)Grok chatdocs.x.ai
DeepSeekDeepSeek-V3 chat and DeepSeek-R1 reasoningapi-docs.deepseek.com
Hugging FaceOpen-weights chat via Inference Providers routerhuggingface.co/docs
AWS BedrockClaude, Llama, Titan, Nova on your AWS accountdocs.aws.amazon.com/bedrock