Trust
Sub-processors
Draft — generated from our provider configuration on 2026-09-23; pending legal review. No row has yet been checked against the executed contract or DPA (“Verified: no”), and entries marked <confirm> are not yet established.
These are the third parties that may process customer data when you use Cortex and the Swfte services. Model providers are configured routes in Swfte's gateway; the table says which ones each intelligence level uses by default. On-device mode sends prompts to none of them.
Providers headquartered in China or Russia are excluded from routing unless a workspace admin opts in to each one. That default takes effect with our compliance release <release date — founder to fill>; until then they remain configured routes that no intelligence level uses by default.
| Sub-processor | Purpose | Location | HQ | Routing | Default for | Verified |
|---|---|---|---|---|---|---|
| Amazon Web Services (AWS)Amazon Web Services EMEA SARL / Amazon Web Services, Inc.Infrastructure | Hosting, databases (DynamoDB), object storage (S3), secrets, logging | EU (Ireland, eu-west-1) — application.yml aws.region default | US | Available | — | No |
| RunPodRunPod, Inc.Infrastructure | GPU compute for self-hosted / serverless open-weight models (when a workspace deploys one) | Region of the provisioned GPU pod | US | Available | — | No |
| OpenAIOpenAI, L.L.C.Model provider | LLM inference (chat, embeddings, image, speech) | United States (EU data residency available per contract) | US | Available | Low, Medium and High; quick tasks at Ultra. | No |
| AnthropicAnthropic PBCModel provider | LLM inference (chat) | United States | US | Available | Ultra (including translation at Ultra). | No |
| Azure OpenAIMicrosoft CorporationModel provider | LLM inference | Azure region of the configured deployment | US | Available | — | No |
| Google (Gemini)Google LLCModel provider | LLM inference (Gemini) | United States / Google Cloud region | US | Available | — | No |
| xAIX.AI LLCModel provider | LLM inference (Grok) | United States | US | Available | — | No |
| Mistral AIMistral AI SASModel provider | LLM inference | European Union | FR | Available | — | No |
| Groq, Inc.Groq, Inc.Model provider | LLM inference (open-weight models) | United States | US | Available | — | No |
| Fireworks AI, Inc.Fireworks AI, Inc.Model provider | LLM inference (open-weight models) | United States | US | Available | — | No |
| Together Computer, Inc.Together Computer, Inc.Model provider | LLM inference (open-weight models) | United States | US | Available | — | No |
| Cohere Inc.Cohere Inc.Model provider | LLM inference, embeddings, rerank | Canada / United States | CA | Available | — | No |
| Perplexity AI, Inc.Perplexity AI, Inc.Model provider | LLM inference with web search | United States | US | Available | — | No |
| Beijing Zhipu Huazhang Technology Co., Ltd.Beijing Zhipu Huazhang Technology Co., Ltd.Model provider | LLM inference (GLM) | People's Republic of China (open.bigmodel.cn) | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| Moonshot AI (Beijing Moonshot AI Technology Co., Ltd.)Moonshot AI (Beijing Moonshot AI Technology Co., Ltd.)Model provider | LLM inference (Kimi) | PRC company; international endpoint api.moonshot.ai | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| Hangzhou DeepSeek Artificial Intelligence Co., Ltd.Hangzhou DeepSeek Artificial Intelligence Co., Ltd.Model provider | LLM inference (DeepSeek) | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| Alibaba Cloud (Alibaba Group)Alibaba Cloud (Alibaba Group)Model provider | LLM inference (Qwen) | PRC company; international endpoint in Singapore (dashscope-intl) | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| SiliconFlowSiliconFlowModel provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| Baichuan IntelligenceBaichuan IntelligenceModel provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| BaiduBaiduModel provider | LLM inference (ERNIE) | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| BaiduBaiduModel provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| BytePlus (ByteDance Ltd.)BytePlus (ByteDance Ltd.)Model provider | LLM inference | ByteDance group; region per endpoint | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| Volcano Engine (ByteDance Ltd.)Volcano Engine (ByteDance Ltd.)Model provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| ByteDance Ltd. (Doubao)ByteDance Ltd. (Doubao)Model provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| ByteDance Ltd. (Coze)ByteDance Ltd. (Coze)Model provider | Agent platform | ByteDance group; region per endpoint | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| iFLYTEK Co.iFLYTEK Co.Model provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| iFLYTEK Co.iFLYTEK Co.Model provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| Ant Group (inclusionAI)Ant Group (inclusionAI)Model provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| Meituan (LongCat)Meituan (LongCat)Model provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| SenseTime Group Inc.SenseTime Group Inc.Model provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| StepFunStepFunModel provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| Tencent Holdings Ltd. (Hunyuan)Tencent Holdings Ltd. (Hunyuan)Model provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| Xiaomi Corporation (MiMo)Xiaomi Corporation (MiMo)Model provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| 01.AI01.AIModel provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| MiniMaxMiniMaxModel provider | LLM inference | People's Republic of China | CN | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| Sberbank (GigaChat)Sberbank (GigaChat)Model provider | LLM inference | Russian Federation | RU | Excluded unless a workspace admin opts in (takes effect with the compliance release) | — | No |
| WorkOSWorkOS, Inc.Business service | Authentication (sign-in via Google, Microsoft or GitHub). | <confirm> | US | — | — | No |
| StripeStripe, Inc. / Stripe Payments Europe, Ltd.Business service | Payment processing for paid plans. | <confirm> | US | — | — | No |
| VercelVercel Inc.Business service | Hosting of the www.swfte.com website. | <confirm> | US | — | — | No |
Changes
Intelligence levels can be re-pointed to other models, so the defaults above can change. How we notify customers of sub-processor changes will be set out in the DPA: <notice period — founder to fill>.