Agent CLI and desktop app · ScaleMax integration

Set up Hermes Agent with the ScaleMax API

Hermes Agent reads custom providers from ~/.hermes/config.yaml. Each model family uses its own API mode — Responses for GPT, Anthropic Messages for Claude, Chat Completions for DeepSeek.

Get an API key Full Hermes Agent guide

Updated October 14, 2026

Choose in the client: custom_providers in ~/.hermes/config.yaml, one per API mode
Protocol: OpenAI Responses, Anthropic Messages, OpenAI Chat Completions
ScaleMax installer: Supported
Video lesson: In the dashboard

Hermes Agent settings for ScaleMax

Enter these values in Hermes Agent. A Pay as you go key and a token pack key use different base URLs — use the one that matches the key you bought.

Setting	Value
Config file	~/.hermes/config.yaml
Key	SCALEMAX_HERMES_API_KEY in ~/.hermes/.env
GPT models	api_mode: codex_responses, base_url https://api.scalemax.pro/v1
Claude models	api_mode: anthropic_messages, base_url https://api.scalemax.pro (token pack: https://api.scalemax.pro/token)
DeepSeek models	api_mode: chat_completions, base_url https://api.scalemax.pro/v1

~/.hermes/config.yaml

custom_providers:
  - name: scalemax-codex
    base_url: https://api.scalemax.pro/v1
    key_env: SCALEMAX_HERMES_API_KEY
    api_mode: codex_responses
    extra_headers:
      User-Agent: ScaleMax-Hermes/1.0

Step by step

Put SCALEMAX_HERMES_API_KEY=your-key in ~/.hermes/.env.
Add one custom provider per model family with the matching api_mode and base URL.
Select it with model.provider: custom:scalemax-codex (or your provider’s name).
Switch models inside a chat with /model custom:scalemax-codex:MODEL_ID.

Or let the installer do it

Choose Hermes. The installer writes the provider blocks and the key file for you.

Terminal

npm install -g scalemax@latest && scalemax

A recorded walkthrough is on the Video setup page in the ScaleMax dashboard (sign-in required).

Models to use in Hermes Agent

Rates are the published ScaleMax rates per million tokens, read from the live catalogue. Your key must include a model before Hermes Agent can call it — GET /v1/models with the key lists exactly which.

GPT models

GPT models for Hermes Agent
Model	Model ID	Input / 1M tokens	Output / 1M tokens
GPT-6.1 Sol	`gpt-6.1-sol`	$2	$10
GPT-6 Sol	`gpt-6-sol`	$2	$10
GPT-6 Luna	`gpt-6-luna`	$0.10	$0.50

Claude models

Claude models for Hermes Agent
Model	Model ID	Input / 1M tokens	Output / 1M tokens
Claude Haiku 4.5	`claude-haiku-4-5-20251001`	$1	$5
Claude Haiku 5.5	`claude-haiku-5-5`	$0.20	$1
Claude Opus 4.6	`claude-opus-4-6`	$5	$25

DeepSeek, Gemini and other models

DeepSeek, Gemini and other models for Hermes Agent
Model	Model ID	Input / 1M tokens	Output / 1M tokens
DeepSeek V4 Pro	`deepseek-v4-pro`	$1	$4
DeepSeek V4 Flash	`deepseek-v4-flash`	$0.19	$0.51
Gemini 3.8 Flash	`gemini-3.8-flash`	$0.75	$3.75

Good to know

Keep the User-Agent header exactly as ScaleMax-Hermes/1.0.
Claude uses the Anthropic origin without /v1; Hermes adds the path itself.
Use DeepSeek models only when their catalogue entry reports tool support.

Troubleshooting

Claude requests fail with a doubled /v1 path: Use https://api.scalemax.pro (no /v1) as base_url for anthropic_messages.
Hermes cannot find the key: Check key_env matches the variable name in ~/.hermes/.env.

ScaleMax is not affiliated with the makers of Hermes Agent. Compatibility describes the API protocol; Hermes Agent's own features and plans are provided by its developer.

Hermes Agent and ScaleMax: questions

Which api_mode should Hermes use for each model?

codex_responses for GPT models on /v1, anthropic_messages for Claude on the API origin without /v1, and chat_completions for DeepSeek on /v1.

Is there a Hermes setup video?

Yes. The Video setup page in the ScaleMax dashboard covers Hermes Agent’s CLI and desktop app on Windows. Sign in to watch it.

Can I put the key directly in config.yaml?

No — keep it in ~/.hermes/.env and reference it with key_env, so the YAML never holds a secret.

Related setup guides

Run Hermes Agent on a prepaid key

Pay by UPI, card or crypto, copy the base URL and model ID, and you are coding in minutes. No subscription.

Get an API key All integrations