Subconscious

How to use Conductor with Subconscious

Let your AI do the setup. Paste this into Claude Code, Cursor, ChatGPT or any assistant.

Set up Conductor to use Subconscious as its model provider.

Subconscious is an inference API for agents. It serves open models through OpenAI- and Anthropic-compatible endpoints, billed per token.
- OpenAI-compatible base URL: https://api.subconscious.dev/v1
- Anthropic-compatible base URL: https://api.subconscious.dev
- Model: subconscious/glm-5.3-marathon
- API key: read it from the SUBCONSCIOUS_API_KEY environment variable. Never commit the key or paste it into project files. If it is not set, ask me to create one at https://platform.subconscious.dev.

Follow these steps. Run commands and edit files yourself. For any step in a settings screen, tell me exactly what to click and what to enter.

1. Create a Subconscious API key
Sign in at https://platform.subconscious.dev and create an API key. You will paste it into the settings screen below.

2. Pick the Claude Code harness
Conductor runs existing agents in isolated workspaces. This setup applies to its Claude Code harness.

3. Set the environment
In Conductor, open Settings, then Environment, and add these variables. ANTHROPIC_API_KEY stays empty so Claude Code does not sign in to Anthropic directly. The Haiku slot drives background tasks, so it gets the cheaper model.
```bash
ANTHROPIC_BASE_URL=https://api.subconscious.dev
ANTHROPIC_AUTH_TOKEN=your_subconscious_key
ANTHROPIC_API_KEY=
ANTHROPIC_MODEL=subconscious/glm-5.3-marathon
ANTHROPIC_DEFAULT_OPUS_MODEL=subconscious/glm-5.3-marathon
ANTHROPIC_DEFAULT_SONNET_MODEL=subconscious/glm-5.3-marathon
ANTHROPIC_DEFAULT_HAIKU_MODEL=subconscious/deepseek-v4.1-flash-marathon
```

4. Start a workspace
Create a workspace and run the agent. Make sure the shell that launches Conductor does not export a real Anthropic key.

Full guide: https://www.subconscious.dev/agents/platforms/conductor.md

When you are done, confirm the tool is using subconscious/glm-5.3-marathon, then summarize what you changed.

To use Conductor with Subconscious, open Settings, then Environment, and set ANTHROPIC_BASE_URL to https://api.subconscious.dev, ANTHROPIC_AUTH_TOKEN to your Subconscious key, and the ANTHROPIC_MODEL variables to subconscious/glm-5.3-marathon. Every Claude Code agent Conductor starts then runs on Subconscious.

Get an API keyRead the API docs

Last verified against the Conductor docs on

Before you start

  • A Mac with Conductor installed
  • A Subconscious account and API key

Conductor docs

Set up Conductor

  1. Step 1: Create a Subconscious API key

    Sign in at https://platform.subconscious.dev and create an API key. You will paste it into the settings screen below.

  2. Step 2: Pick the Claude Code harness

    Conductor runs existing agents in isolated workspaces. This setup applies to its Claude Code harness.

  3. Step 3: Set the environment

    In Conductor, open Settings, then Environment, and add these variables. ANTHROPIC_API_KEY stays empty so Claude Code does not sign in to Anthropic directly. The Haiku slot drives background tasks, so it gets the cheaper model.

    Settings → Environment
    ANTHROPIC_BASE_URL=https://api.subconscious.dev
ANTHROPIC_AUTH_TOKEN=your_subconscious_key
ANTHROPIC_API_KEY=
ANTHROPIC_MODEL=subconscious/glm-5.3-marathon
ANTHROPIC_DEFAULT_OPUS_MODEL=subconscious/glm-5.3-marathon
ANTHROPIC_DEFAULT_SONNET_MODEL=subconscious/glm-5.3-marathon
ANTHROPIC_DEFAULT_HAIKU_MODEL=subconscious/deepseek-v4.1-flash-marathon

  4. Step 4: Start a workspace

    Create a workspace and run the agent. Make sure the shell that launches Conductor does not export a real Anthropic key.

Why run Conductor on Subconscious

Lower cost on long tasks
50% to 80% lower cost on long tasks, because the runtime processes far fewer tokens. Cached input bills at a steep discount, and agentic coding sees cache hit rates above 95%.
One plan for the whole team
Token plans give a fixed monthly price for a daily token allowance, shared across models and across your team. Enterprise deployments run inside your own cloud.
Long sessions stay fast
2x faster task completion on long tasks than standard open-model inference, especially past 200k tokens of context.
Your data stays yours
Prompts, inputs and outputs are not recorded, and nothing you send is used to train a model.

Connection details

For any client that takes a custom OpenAI- or Anthropic-compatible endpoint.

OpenAI-compatible base URL
https://api.subconscious.dev/v1
Anthropic-compatible base URL
https://api.subconscious.dev
API key
From subc login or the Subconscious dashboard
Model
subconscious/glm-5.3-marathon
Multimodal, high-throughput model
subconscious/deepseek-v4.1-flash-marathon

Full reference in the Subconscious API docs.

Questions

Does Conductor charge for model usage?
No. Conductor orchestrates agents such as Claude Code and Codex and does not resell model usage, so tokens bill to your Subconscious account.
Which Subconscious model should I use with Conductor?
Start with subconscious/glm-5.3-marathon, an open frontier coding model built for long agent runs. subconscious/deepseek-v4.1-flash-marathon is a cheaper, high-throughput alternative that also takes image input. Use the full model ID, including the subconscious/ prefix.
Is Conductor on Subconscious OpenAI- or Anthropic-compatible?
This guide connects Conductor through the Anthropic-compatible Subconscious API. Subconscious serves both formats: OpenAI Chat Completions at https://api.subconscious.dev/v1 and Anthropic Messages at https://api.subconscious.dev.
How is Conductor usage on Subconscious billed?
Per token by default, with no commitment. Heavy users can switch to a monthly token plan: a fixed price for a large daily token allowance shared across your team.

More agent platforms

See all