Are all models US-hosted?

[pure stump]

(I am a paid user) Models like Kimi and GLM used to say "us-hosted". I don't see that anymore. Are ALL models US-hosted, or does it vary? I'm particularly interested in if minimax is US-hosted, but really my question applies to all models. I need to be able to trust that none of my data is going to a server in another country or that changes my ability to use those other models. Thank you!

[jolly widget]

Hi @pure stump. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:

---

Great question, and I understand why data locality matters to you. Here's what the knowledge sources directly support:

Fireworks AI-hosted models (Kimi, GLM, Minimax, Qwen)

These open-source models — including Minimax 2.7, GLM 5/5.1, Kimi K2.5/K2.6, and Qwen 3.6 Plus — are hosted via Fireworks AI. According to Warp's subprocessors list, Fireworks AI is listed as a US-based subprocessor used for "AI enabled functionality." [Subprocessors]

Other model providers (Anthropic, OpenAI, Google)

These are also listed as US-based subprocessors on the same page. [Subprocessors]

Data retention

For Enterprise users, Warp maintains Zero Data Retention (ZDR) agreements with its contracted LLM providers (Anthropic, OpenAI, Google), meaning they do not store or train on your data. [Trust Center]

Important caveat

The knowledge sources confirm that all listed subprocessors are US-located, which covers all the model providers (Fireworks AI, Anthropic, OpenAI, Google). However, the knowledge sources do not explicitly state that every individual model's inference compute runs exclusively on US soil — only that the subprocessors themselves are US-based. If you need a contractual guarantee about data residency at the infrastructure level, I'd recommend reaching out directly to [email protected] for confirmation. [Security Overview]

Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:

@pure stump Kapa.ai is still learning and improving, please let me know how I did by reacting below.