Model-Agnostic AI Orchestration

We are model and token agnostic. We orchestrate your AI workflows using your chosen AI — whether free, cheap, or bleeding edge. Whether on-premise or in the cloud, private models or public APIs. We make it simple.

Our Philosophy

UpGPT is a workflow orchestration platform, not an AI provider. We believe you should choose the AI that works for your business — not be locked into ours. Connect your own API keys for any supported provider, giving you full control over model selection, cost, and data privacy. We handle the orchestration, routing, and tool integration. You decide which models power your workflows. We also offer hosted compute for smaller customers just getting started — so you can begin immediately and bring your own keys as you scale.

Why Model-Agnostic?

Cost Control

Use free models (Gemma, Llama) for routine tasks, premium models (GPT-4o, Claude) only when needed. You set the budget.

Privacy

Your data goes directly to your chosen provider. UpGPT never sees your prompts or responses.

No Vendor Lock-in

Switch models anytime. If a cheaper or better model launches tomorrow, use it immediately.

Compliance

Use on-premise models for sensitive data. Run Gemma locally with zero cloud exposure.

Supported Providers

OpenAI

GPT-4o, GPT-4o-mini

Industry-standard models for general-purpose tasks.

Anthropic

Claude Sonnet, Claude Opus, Claude Haiku

Strong reasoning and long-context capabilities.

Google

Gemini Pro, Gemini Flash

Fast, multimodal models with generous context windows.

OpenRouter

100+ models, single API key

Universal gateway to every major model provider. One key, all models.

Local Models

Ollama, LM Studio

Run models on your own hardware via the UpLink desktop app. Zero cloud exposure.

Custom Endpoints

Any OpenAI-compatible API

Connect any service that implements the OpenAI chat completions format.

How It Works

1

Add Your API Key

Go to Settings → API Keys in your UpGPT dashboard. Paste your provider's API key and select the provider type.

2

Choose Your Models

Select which models to use for different task types. Assign fast models to simple tasks and powerful models to complex ones.

3

UpGPT Routes Automatically

UpGPT analyzes each task and routes it to the right model based on complexity. No manual intervention needed.

4

Monitor Usage

Track per-task cost directly in your provider's dashboard. UpGPT never charges for AI usage -- you pay your providers directly.

How AI Works Across Plans

PlanAI Model AccessDetails
FreeYour KeysConnect your own AI provider from day one. Full model choice, zero platform markup.
Starter & Solo ProHosted AI IncludedWe handle the AI so you can get started fast. Switch to your own keys anytime for more control.
Growth & aboveYour KeysAs you scale, you bring your own keys. Full control over models, costs, and data residency.

Ready to connect your AI providers?