Flash as the primary model

Hermes — direct connection

Settings checked against Hermes 0.21.0. Install Hermes using its official guide and create a separate profile. No private wrapper, GPU or server Docker image is required.

hermes --version
hermes profile create flash
hermes -p flash config edit

# Teralor Flash as the PRIMARY model in Hermes 0.21 (verified against 0.21.0).
# Apply to a NEW profile so existing settings are never overwritten:
#   hermes profile create flash
#   hermes -p flash config edit      # merge this YAML, do not replace other settings
#   hermes -p flash chat
# The key is read from the TERALOR_API_KEY environment variable — never paste a
# key literal into this file, a prompt, or source control.
providers:
  teralor-flash:
    api: https://api.teralor.com/v1
    key_env: TERALOR_API_KEY
    transport: chat_completions
    default_model: Flash
    discover_models: true
    models:
      Flash:
        context_length: 262144
    extra_body:
      max_tokens: 32768
model:
  provider: custom:teralor-flash
  default: Flash
context:
  engine: compressor
compression:
  enabled: true
  threshold_tokens: 172032

hermes -p flash config get providers.teralor-flash.models.Flash.context_length
hermes -p flash config get compression.threshold_tokens
hermes -p flash chat

Pi — documented setup

Merge the provider into ~/.pi/agent/models.json and the compaction settings into .pi/settings.json in your Flash project. Restart Pi, open /model and select teralor-flash/Flash. These examples follow the Pi model documentation; Pi was not locally exercised in our audit.

{
  "providers": {
    "teralor-flash": {
      "baseUrl": "https://api.teralor.com/v1",
      "api": "openai-completions",
      "apiKey": "$TERALOR_API_KEY",
      "authHeader": true,
      "models": [
        {
          "id": "Flash",
          "name": "Teralor Flash",
          "input": ["text"],
          "reasoning": false,
          "contextWindow": 262144,
          "maxTokens": 32768,
          "compat": {
            "maxTokensField": "max_tokens",
            "supportsDeveloperRole": false,
            "supportsReasoningEffort": false,
            "supportsStore": false
          }
        }
      ]
    }
  }
}

{
  "compaction": {
    "enabled": true,
    "reserveTokens": 90112,
    "keepRecentTokens": 20000
  }
}

ZCode — documented setup

Follow the provider steps below, then select Flash as the chat model. These GUI instructions follow the linked ZCode docs and are not a claim of live Flash certification.

1. Install ZCode from its official docs, then open the chat model selector → Manage Models → Settings / Model Settings (the docs also call this Model Providers). See the ZCode configuration docs.
2. Add a Provider named Teralor Flash; choose the OpenAI-compatible protocol/endpoint, not Anthropic. The API base URL must be https://api.teralor.com/v1(match the environment), and enter your own matching customer key in the app's API-key field — never paste the key into chat or source control.
3. Add a Model with the exact ID Flash and enable the provider/model.
4. Open the model's Advanced settings: context window 262144, Max output tokens 32768, or lower if your authenticated /v1/models result says so. These are the product context and recommended per-reply output, not additive limits. Custom-provider context edits apply to new sessions after saving.
5. Keep optional thinking controls unset unless your route supports them. Provider options only accept connection fields — arbitrary reasoning/extra_body keys are silently ignored.

6. Save, start a new task/session, select Flash in the model picker, and ask a bounded read-only tool task. Inspect the result before allowing edits.

ZCode compacts automatically with output and safety headroom and exposes no user-facing compaction threshold — do not add the 172032 trigger or extra JSON request fields here.

Cursor — BYOK, with limitations

1. In Cursor Settings → Models → API Keys, enter your customer key in the OpenAI key field, enable Use OpenAI API Key, enable Override OpenAI Base URL and set https://api.teralor.com/v1. See the Cursor API-keys docs.
2. Add a custom model with the exact server ID Flash, enable it and select it explicitly (not Auto), then start a new Agent chat with a short read-only test. Availability can depend on your account/plan.

This is a docs/support-backed BYOK recipe, not a tested native Flash integration. Unlike Hermes, Pi or ZCode, Cursor cannot detect or manually set the true window for unknown custom models; support reports a 1M assumption for unknown IDs, so it can hit Flash's actual 262144 limit before auto-compaction. The cited guidance provides no per-custom-model input/output/compaction controls to configure here, and Max Mode is not a verified fix. Keep tasks and context small, start fresh chats, and use executor mode when you need explicit budgets or long tasks.

Note: the base-URL override applies broadly to OpenAI-family requests, not only Flash — turn it off when returning to normal Cursor/OpenAI models. Your key and requests transit Cursor's backend, and BYOK does not apply to Tab completion.

Codex — executor only

Using Flash as the orchestrator in Codex is not possible at the moment. We recommend using Flash only as an executor in Codex — see executor mode.

Claude Code — executor only

Using Flash as the orchestrator in Claude Code is not possible at the moment. We recommend using Flash only as an executor in Claude Code — see executor mode.