Filter by where providers report running inference. A provider may operate in several regions; these research filters do not restrict your API traffic. Manage your routing policy →
Provider Headquarters Runs in Prompt retention Trains on API data
OpenAI 🇺🇸 United States US, EU, Other 30 days no ▼ Google Vertex AI 🇺🇸 United States US, EU, Other up to 24h (cache); up to 30 days (flagged abuse logs) no ▼ xAI 🇺🇸 United States US 30 days no ▼ Meta 🇺🇸 United States undisclosed undisclosed no ▼ Mistral 🇫🇷 France EU, US 30 days undisclosed ▼ Z.AI 🇨🇳 China Other none no ▼ MiniMax 🇨🇳 China undisclosed yes undisclosed ▼ Baidu 🇨🇳 China undisclosed yes yes ▼ Tencent 🇨🇳 China undisclosed undisclosed undisclosed ▼ Azure 🇺🇸 United States US, EU, Other yes no ▼ Cohere 🇨🇦 Canada undisclosed 30 days yes ▼ Groq 🇺🇸 United States US, EU, Other none no ▼ Fireworks 🇺🇸 United States US, EU, Other none no ▼ Nebius 🇳🇱 Netherlands EU, US, Other no fixed window (inference cache) no ▼ SambaNova 🇺🇸 United States undisclosed none undisclosed ▼ NVIDIA 🇺🇸 United States undisclosed none yes ▼ Databricks 🇺🇸 United States US, EU, Other up to 30 days no ▼ AI21 🇮🇱 Israel undisclosed undisclosed no ▼ AionLabs 🇵🇱 Poland undisclosed undisclosed undisclosed ▼ AkashML 🇺🇸 United States undisclosed none no ▼ Ambient 🇺🇸 United States undisclosed none no ▼ Arcee AI 🇺🇸 United States undisclosed 30 days no ▼ Atlas Cloud 🇺🇸 United States US up to 7 days (standard) no ▼ Avian 🇺🇸 United States US none no ▼ Baseten 🇺🇸 United States undisclosed none undisclosed ▼ Chutes 🇰🇳 Saint Kitts and Nevis undisclosed none no ▼ Cirrascale 🇺🇸 United States US undisclosed undisclosed ▼ Clarifai 🇺🇸 United States undisclosed until deleted by customer no ▼ CoreWeave 🇺🇸 United States undisclosed undisclosed undisclosed ▼ Cosine 🇬🇧 United Kingdom undisclosed undisclosed undisclosed ▼ C Crucible undisclosed undisclosed undisclosed undisclosed ▼ Crusoe 🇺🇸 United States undisclosed none no ▼ Darkbloom 🇺🇸 United States US, Other undisclosed no ▼ Friendli 🇺🇸 United States US until organization/project deletion + up to 1 day undisclosed ▼ GMI Cloud 🇺🇸 United States undisclosed undisclosed undisclosed ▼ Inferact 🇺🇸 United States undisclosed undisclosed undisclosed ▼ Inflection 🇺🇸 United States undisclosed undisclosed undisclosed ▼ io.net 🇺🇸 United States undisclosed undisclosed undisclosed ▼ ionstream 🇺🇸 United States undisclosed undisclosed undisclosed ▼ Liquid 🇺🇸 United States US yes yes ▼ Makora 🇺🇸 United States US 30 days (Starter/Developer tiers); transient-only on Developer Pro/Pay-as-you-go no ▼ Mancer 🇺🇸 United States US none no ▼ Mara 🇺🇸 United States undisclosed none no ▼ Modal 🇺🇸 United States US, EU, Other undisclosed undisclosed ▼ ModelRun 🇺🇸 United States US undisclosed undisclosed ▼ Modular 🇺🇸 United States US undisclosed undisclosed ▼ Morph 🇺🇸 United States undisclosed none no ▼ NCompass 🇺🇸 United States undisclosed undisclosed undisclosed ▼ Nex AGI 🇨🇳 China undisclosed undisclosed undisclosed ▼ NextBit 🇪🇸 Spain EU, Other up to 90 days no ▼ Novita undisclosed undisclosed none no ▼ Ollama 🇺🇸 United States undisclosed none no ▼ Parasail 🇺🇸 United States undisclosed none no ▼ Perceptron 🇺🇸 United States undisclosed undisclosed undisclosed ▼ Phala 🇺🇸 United States undisclosed none undisclosed ▼ Poolside 🇺🇸 United States US, Other not disclosed yes ▼ Q Quiver undisclosed undisclosed undisclosed undisclosed ▼ Reka 🇺🇸 United States undisclosed yes no ▼ Relace 🇺🇸 United States undisclosed not disclosed yes ▼ Seed 🇸🇬 Singapore undisclosed undisclosed undisclosed ▼ Sourceful 🇬🇧 United Kingdom undisclosed undisclosed undisclosed ▼ StreamLake 🇸🇬 Singapore undisclosed not disclosed yes ▼ S Switchpoint undisclosed undisclosed undisclosed undisclosed ▼ Tenstorrent 🇺🇸 United States undisclosed undisclosed undisclosed ▼ Upstage 🇰🇷 South Korea undisclosed none no ▼ Venice 🇺🇸 United States US none no ▼ Wafer 🇺🇸 United States undisclosed 30 days (non-ZDR requests) no ▼ Xiaomi 🇨🇳 China EU, Other not disclosed yes ▼