BYOK - Bring Your Own Keys

Last updated: April 9, 2026


BYOK (Bring Your Own Keys) allows your organization to connect Scenario directly to your own API accounts at supported third-party model providers. When enabled, generations using those specific models are routed through your personal API key - billed directly to your provider account - instead of consuming Scenario credits.

Why Use BYOK?

This gives enterprise teams two significant advantages:

  • Cost Optimization: Use models from providers where you already have negotiated pricing or committed spend.

  • Billing Transparency: Usage appears in your own provider dashboard, making it straightforward to track and report AI spend alongside your other operations.

Supported Providers

The Scenario platform supports an expanded list of AI providers for BYOK integration:

  • OpenAI: GPT Image and other DALL-E based models.

  • Anthropic: Claude models for LLM and workflow nodes.

  • xAI: Grok-based generation and editing models.

  • Bytedance: MiniMax audio and video models.

  • Tencent: Hunyuan image and video models.

  • GCP (Google VertexAI): Gemini and Imagen model families.

  • Tripo AI: Specialized 3D generation models.

Note: Additional providers may be added in future releases. Check the Organization Settings page for the most current list.

Post-processing Fees

While your API key covers the generation cost (billed directly by your provider), Scenario applies a separate charge for platform-side operations like previews, rendering, and indexing.

Post-processing Rates

(Effective as of March 2026)

Asset Type

Scenario Fee (per asset)

🖼 Image

1 CU

🎬 Video

5 CUs

🎵 Audio

2 CUs

🧊 3D

3 CUs

📝 Text

0.5 CUs

Why this fee? Even when using your own keys, Scenario's infrastructure handles the heavy lifting of processing, storing, and displaying the generated assets within your workspace.

How Credits Work with BYOK

When BYOK is active for a specific model:

  1. Generation Cost: Billed to your provider account (OpenAI, Google, etc.) at their standard rates.

  2. Platform Fee: Scenario deducts the applicable Post-processing CU fee (listed above) from your Scenario credit balance to handle the rendering and asset management.

  3. Native Features: Training, storage, and Scenario-native models continue to use standard platform credits.

Setting Up BYOK

BYOK is configured at the organization level by an admin. These settings apply organization-wide.

  1. Open Organization Settings: Navigate to Organization Settings from the main menu and select the External Keys section.

  2. Select the provider you want to configure and enter your API key. You can obtain your keys at the following official developer portals:

  3. Confirm the Connection: Scenario will validate the key. Ensure it has generation permissions enabled.

  4. Select Models: Choose which specific models should route through BYOK and which should remain on the Scenario credit system.

  5. Notify Your Team: Routing happens automatically for members. Let them know which models are BYOK-enabled so they understand where usage charges will appear.

What BYOK Does Not Cover

BYOK applies only to the third-party generation cost. It does not cover:

  • Scenario Platform Services: Training, workflow infrastructure, storage, and other native services still require a Scenario subscription or credit balance.

  • Native Models: Models unique to Scenario with no third-party equivalent.

  • Provider Limits: If your OpenAI/Google account is rate-limited or suspended, the generations will fail regardless of your Scenario status.

Team Model Allowlist

Introduced alongside BYOK, the Team Model Allowlist allows admins to restrict which models are available to team members. This is ideal for cost control, compliance, or content policy reasons.

  • Found in Organization Settings > Models.

  • Admins can still see all models; non-admin members only see the "Allowed" list.

  • BYOK and the Allowlist work independently - you can use one without the other.

Security & Transparency

  • Encrypted Storage: Your keys are never exposed in the UI after the initial setup.

  • Usage Tracking: Generation spend appears in your provider's dashboard, while post-processing fees appear in your Scenario Usage reports.

  • Rate Limits: Generations remain subject to the rate limits and quotas of your specific provider account.