How to use MiniMax in Factory

MiniMax is a family of open-weight language models from MiniMax. Factory offers MiniMax natively, so you can run it inside a Droid without provisioning your own inference or bringing a key. If you have been searching for where to run MiniMax for coding, this is how it fits into Factory.

MiniMax appears in Factory's model list next to the Anthropic, OpenAI, Google, and xAI models. Selection works the same as for any other model, and Factory serves the inference and handles billing.

What MiniMax is

MiniMax is an open-weight model series, meaning the weights are published rather than served only through a single vendor's API. Factory groups it under Droid Core, the open models it hosts and serves directly. Three variants are available today: MiniMax M3, M2.7, and M2.5, all listed on the Available Models page with their IDs and settings.

Because Factory serves these models, MiniMax works across the surfaces where Droids run, with no serving infrastructure to manage on your side.

Selecting MiniMax in a Droid

In an interactive session, run /model and pick the MiniMax variant you want, or cycle models with Ctrl+N from the chat input. The interactive commands are documented in the Droid CLI reference.

For headless runs, pass the model ID with -m:

droid exec -m minimax-m3 --auto medium "add pagination to the results endpoint and cover it with tests"

MiniMax M3 and M2.7 run at High reasoning; M2.5 exposes Low, Medium, and High (the default). Set the effort with Tab in interactive mode or -r in droid exec. The published levels are on the models page.

Cost and the credit multiplier

Factory bills usage with a per-model credit multiplier, and MiniMax carries the lowest multiplier in the current catalog. All three variants are listed at 0.12x, which makes MiniMax attractive for high-volume, cost-sensitive work. Verify the live multiplier on the Available Models page before you standardize on it, since these values change.

The table below summarizes the MiniMax options as published.

When to reach for MiniMax

The low multiplier makes MiniMax a candidate for large batches of scoped changes, cleanup passes, or experimentation where you want to keep credit spend down. Keeping model choice open is the case for a model-agnostic setup, and Factory lets you switch models per task. When you would rather have Factory choose the model for each step, Factory Router routes across the catalog for you.

MiniMax in headless runs and Missions

The same model IDs work in CI and automation. A GitHub Actions step can run droid exec -m minimax-m3 to keep automated tasks on the lowest-multiplier model. In Missions, assign MiniMax to worker agents with --worker-model minimax-m3 while keeping a stronger model for validation, so you concentrate credits on the steps that need reasoning.

Further reading

• How to use GLM-5.2 in Factory
• How to use DeepSeek V4 in Factory
• Why your AI agents should be model-agnostic
• Available Models