Cloud Hub // Every popular provider

Use the model you want. Keep the Celestos interface.

Cloud Hub connects Celestos to the popular model providers people actually use. Local stays the default, but frontier cloud models are ready when you need more context, faster throughput, stronger reasoning, or a specific model family.

See Models Get Started

Providers // Built-in catalog

OpenAI, Anthropic, Google, OpenRouter, Groq, and more.

Provider group	What it is good for
OpenAI, Anthropic, Google AI	Frontier reasoning, coding, vision, writing, and general assistants.
OpenRouter, Together AI, Fireworks	Broad open-model catalogs, routing flexibility, and availability.
Groq, Cerebras, SambaNova	Ultra-fast inference and free-tier experimentation.
Hugging Face, NVIDIA NIM, Cohere	Serverless models, enterprise endpoints, RAG, and structured output.
DeepSeek, Mistral AI, xAI, Perplexity	Low-cost reasoning, coding, long context, and web-grounded answers.
Kie.ai, CLōD	Aggregator and curated endpoint options for cost, access, and availability.

What Cloud Hub does

Loads live model catalogs instead of stale dropdowns

Shows ready status for free tiers and configured API keys

Links straight to provider key setup when a key is missing

Sends a remote model to Chat with one click

Lets you hide noisy models so catalogs stay usable

Routing // Local first, cloud when useful

Cloud is not the product. It is the extra gear.

Celestos keeps privacy-sensitive work on local models by default. Switch a role, workflow, or one-off request to cloud when the task needs frontier reasoning, a huge context window, real-time web grounding, or faster inference.

Private local default

No API cost

Run everyday chat, documents, and sensitive work on your machine.

Frontier role upgrade

For hard thinking

Send only the manager, judge, or hardest step of a Hive run to a frontier provider.

Provider flexibility

No lock-in

Move between providers based on speed, cost, model access, or task type.