Atlas Cloud includes an optional model router: a hosted service for model routing, provider policy, credit checks, rate limits, and usage events. It sits behind an Atlas runtime feature flag, so bring-your-own-keys stays the default and you opt in only when a team wants centrally managed routing instead of per-developer provider keys.
Why a hosted router at all
Bring-your-own-keys is ideal for individuals, but teams often want one place to set provider policy, enforce rate limits, and account for spend. The model router is that place, offered as an opt-in service rather than a requirement.
With hosted routing, an organization can decide which providers and models are allowed, apply credit checks before a run, and collect usage events for reporting, without distributing raw provider keys to every seat. Developers still work in the same terminal agent; only the routing decision moves.
Still your choice, per run
The router is gated behind an Atlas runtime feature flag. Nothing about your existing local setup changes unless you turn it on, and you can keep routing directly to your own keys or local Ollama models.
That means the privacy story is intact: local and BYO-key routing remain first-class. Hosted routing is an added option for teams that value central control and reporting, not a toll you must pay to use Atlas.
Usage without exposing content
The router records usage events (which model, how much) for credit and rate-limit decisions. It is part of the control plane, which is designed not to require prompts, model outputs, or provider request bodies for ordinary operations.
So a team can meter and cap usage while the substance of each request, your code and prompts, stays out of the accounting path. You get spend visibility and policy enforcement without turning the router into a content log.
Frequently asked questions
- Does hosted routing replace bring-your-own-keys?
- No. Bring-your-own-keys and local models are the default. Hosted routing is opt-in behind a runtime feature flag for teams that want centralized provider policy and usage control.
- What does the model router manage?
- Model routing, provider policy, credit checks, rate limits, and usage events, so teams can standardize and meter model access.
- Does the router see my prompts or code?
- The router is part of the control plane, which is designed not to require prompts, model outputs, or provider request bodies for ordinary operations. It records usage events, not content.
- How do I turn it on?
- Hosted routing is gated behind an Atlas runtime feature flag, so it is off until you explicitly enable it.
Try Atlas in your terminal
The terminal-native AI coding agent. Free core, single binary.
Install AtlasRelated guides
Atlas for Erlang in 2026
Atlas is a terminal-native AI coding agent for Erlang/OTP in 2026. Run it in an app with a rebar.config, map supervisors and gen_server modules, review every diff.
Atlas vs CodeGPT in 2026: A Developer's Guide to Terminal and IDE AI Agents
Comparing Atlas and CodeGPT in 2026 for developers. Atlas offers terminal-native TUI and permission-gated tools, while CodeGPT provides IDE integration and a full repo Knowledge Graph.
Atlas with Mistral Medium 3.1 (2508): Setup, Cost, and Fit in 2026
Run Atlas on Mistral Medium 3.1 (2508): a 262,144 token window at $0.40 / 1M input tokens and $2.00 / 1M output tokens. Setup steps, costs, and honest tradeoffs.
Atlas with Llama 3.2 1B (local via Ollama): the 2026 offline small_model slot
Llama 3.2 1B (local via Ollama) in Atlas: a 1.3GB CPU-friendly pull with a 128,000 token window, Free (self-hosted), best wired to small_model, not model.
Atlas with Command R7B in 2026: Cohere's Cheapest Model, Used Correctly
Command R7B costs $0.0375 per Mtok input, about 1/66th of Command A, and still carries a 128,000 token context. Here is how to slot it into Atlas without wrecking your code.
Atlas with GPT-5.1 Codex: Plan First, Then Build in 2026
GPT-5.1 Codex in Atlas: a frontier model from OpenAI at $1.25 per Mtok input, $10 per Mtok output, built to execute an approved plan across a long tool chain.
Atlas vs Sourcegraph Cody: Terminal AI Coding Agents in 2026
Comparing Atlas, the terminal-native AI coding agent, with Sourcegraph Cody, an editor extension for whole-codebase search, for developers in 2026. Explore features, pricing, and workflow.
Atlas with Mistral Medium 3.5: The EU-Hosted Frontier Model (2026)
Mistral Medium 3.5 drives Atlas at $1.50 / $7.50 per Mtok on a 262,144 token window with a matching 262,144 output limit. The EU-hosted option for data residency.