Wirestack
Cohere Model Vault now runs open models like DeepSeek and Gemma, in US, EU or Canada regions (November 3, 2026)
Cohere's dedicated Model Vault now hosts open models from DeepSeek, Google, OpenAI and others, with EU and Canada regions.
Your keys
0 of 1 ready
Create each key on its own page, paste it below, then copy the finished lines into your project's .env file. What you paste stays in this tab. It is never sent to Wirestack or saved.
- Cohere API keyCO_API_KEYGet key
Sign in to the Cohere dashboard, open the API keys page and create or copy a key. Trial keys are free.
3 of 3 free copies left this month.Unlimited with Pro
What changed
On November 3, 2026 Cohere opened Model Vault to open models from other labs. Model Vault is Cohere's managed, dedicated hosting: a vault runs models on GPUs reserved for you, and Cohere now serves open models in it next to its own Command, Embed, Rerank and other models.
- Standard Model Vault runs open models from DeepSeek, Z.ai (GLM), Xiaomi (MiMo), NVIDIA (Nemotron), Google (Gemma), OpenAI (GPT-OSS), Microsoft (Phi), Alibaba (Qwen) and Mistral AI. The blog post names DeepSeek V4.1 Flash, GLM 5.3 Flash, MiMo V2.6 Pro, Gemma 4, NVIDIA Nemotron, GPT-OSS, Phi and Hy3.
- Model Vault Encrypted runs Gemma, GPT-OSS, Phi, Qwen, Devstral and Ministral with confidential computing, so neither Cohere nor the cloud provider can see prompts or responses.
- Regions: Standard vaults can be created in the United States, the European Union or Canada. Encrypted vaults are US only for now; EU and Canada are listed as coming soon. A vault's region can't be changed after you create it.
- Zero Data Retention is the default for Model Vault: prompts and responses are not kept.
- Pricing is per GPU, per hour, for both vault types. Reserved capacity goes through Cohere sales.
- Open models use the OpenAI-compatible Chat Completions API at your vault's endpoint.
Prerequisites
- A Cohere API key in
CO_API_KEY - A vault created in the Model Vault app at vault.cohere.com, and its endpoint URL
- Python with the OpenAI SDK (
pip install openai)
Steps
- Sign in at vault.cohere.com and create a vault. Pick the region first: the EU and Canada are only offered for Standard vaults, and the region is fixed once the vault exists.
- Add an open model to the vault and copy the vault's endpoint URL.
- Call it with the OpenAI SDK. This is Cohere's example; replace
<COHERE_API_KEY>with your key and<YOUR_VAULT_ENDPOINT_URL>with your endpoint:
from openai import OpenAI
client = OpenAI(
api_key="<COHERE_API_KEY>",
base_url="<YOUR_VAULT_ENDPOINT_URL>/v1",
)
response = client.chat.completions.create(
model="deepseek-v4-1-flash",
messages=[{"role": "user", "content": "Hello from Model Vault!"}],
)
print(response.choices[0].message.content)