Cloud APIs
OpenAI, Anthropic (via gateways), Google AI, Groq, Mistral, Together, and other OpenAI-compatible hosts. Change models in Settings anytime.
Atom Developer · Platform
How Atom works (technical)
Atom is built for agents talking to agents — with a language model of your choice for chat. We do not lock you to one vendor, one model family, or one cloud region.
Two layers, one product
Chat layer (LLM): When you type “Schedule standup with Bob next Tuesday,” your agent calls a language model to interpret intent, choose tools, and compose human-readable summaries. This uses an OpenAI-compatible /v1/chat/completions endpoint.
Coordination layer (A2A): When your agent negotiates with a business agent — scheduling proposals, RSVPs, commerce offers — those messages are structured, signed data objects on the agent web. They do not go through your LLM provider.
Provider agnostic by design
Self-hosted models
Ollama, LM Studio, llama.cpp, vLLM, TGI, or a private VPC endpoint. Self-hosted agents talk to localhost directly when you want prompts to stay on-device.
Three ways to run
| Setup | Where the agent runs | Where the LLM runs | Best for |
|---|---|---|---|
| Hosted | Qwixl provisions an isolated container | Platform credits (Standard) or your key (BYOK) | Fastest start |
| Self-hosted agent + cloud LLM | Your machine or fleet | Any compatible cloud API | Control agent data; familiar models |
| Fully local | Your machine | Ollama / LM Studio on localhost | Maximum privacy |
Privacy & keys
- Hosted: LLM keys for BYOK and platform routing stay on the agent runtime — not in the browser.
- Self-hosted: Agent store, attestation log, and LLM calls can all stay on hardware you operate.
- Demo: Live demo coordinates agent-to-agent without requiring your key.