Choosing AI Models

Each provider in Provider Mode offers its own lineup of models, and the model you pick shapes quality, speed, and cost. Corkulous shows the models supported by each provider integration — including provider-specific media models where available — so you can choose separately for text, images, audio, music, and video.

Text Generation Models

Choose the model used for board generation and the Agent. Each provider offers a range of models with different strengths:

Provider	Strengths
OpenAI	Supported GPT text and reasoning models for board generation and Agent requests. Use the current model picker for available IDs.
Anthropic	Excellent reasoning, nuanced writing, and long document handling.
Google	Good all-around with fast and capable options. Supports multimodal input.
xAI	Grok text models and supported web tools. Current information requires an enabled retrieval/search capability; a model alone is not a live source.

Models are listed in the Providers tab based on what each provider integration supports. More capable models generally cost more per use.

OpenAI Tuning Note

Sampling and reasoning controls depend on the selected model. Some models restrict temperature or topP when reasoning is enabled. Use the controls available in the provider’s Advanced section; Corkulous validates them against its supported model metadata.

Image Models

Provider	Description
OpenAI	Premium image generation with strong prompt following and option-aware pricing for supported models.
Google	Imagen models for high-quality image generation.
xAI	Image generation through Grok models.

Select your preferred image provider in the General tab and choose a specific model in the Providers tab. Then use that provider card’s Defaults section to save image-specific defaults such as size, aspect ratio, quality, style, or count.

Audio and Music Models

Provider	Features
OpenAI	Natural-sounding narration voices, fast generation.
ElevenLabs	Provider Mode music generation and premium speech voices with more variety and expressiveness.

Configure OpenAI and ElevenLabs audio or music models in the Providers tab. Use each provider card’s Defaults section to save the voice, output format, or music defaults you want reused later.

Video Models

Provider	Description
Google	Veo models with multiple style options.
xAI	Video generation through Grok models.

Select your preferred video provider in the General tab and choose a specific model in the Providers tab. Then use that provider card’s Defaults section to save duration, aspect ratio, resolution, and audio preferences.

How to Select Models

1. Go to Settings → AI Settings.
2. Open the Providers tab.
3. Under each provider, tap the model picker for text, image, audio, or video.
4. Select your preferred model from the list.
5. Open the provider’s Defaults section if you want to save provider-specific media defaults for that model.
6. Open the provider’s Advanced section if you want to adjust reasoning, timeout, temperature, top-p, or max output tokens for that provider.

Switching Models

When you switch models, Corkulous re-checks the saved defaults against the new model’s capabilities. Unsupported sizes, voices, resolutions, or other options are repaired automatically, so future generation views and worker runs use valid settings.

Availability and Cost

Models are only available if you have added an API key for that provider and the provider is enabled. More capable models and higher-quality media options generally cost more. Check your provider dashboard for current pricing when using Provider Mode; Credits mode shows Corkulous credit estimates before generation.

Known retired models may be replaced automatically with a supported option when one is available. Review any repair warning in settings and confirm the effective provider/model before generating. Unknown model IDs are still invalid.