Skip to main content

​
Overview

Whim supports three AI providers, each running natively in task containers. You can switch providers per workspace or per task — every provider has access to the same dev environment, terminal, and git tools.
CCR + OpenRouterClaude SubscriptionCodex Subscription
RuntimeCloud Code Runtime (CCR)Native Claude CLINative Codex CLI
AuthManaged by WhimYour Anthropic subscriptionYour ChatGPT account
Default modelClaude Sonnet 4.5Claude Opus 4.6GPT 5.4
Available models20+ (Claude, GPT, Gemini, Grok, DeepSeek, and more)Claude family (Opus, Sonnet, Haiku)GPT 5.x family
Setup requiredNoneOAuth tokenChatGPT login
CU costContainer time + API tokensContainer time onlyContainer time only
Fast modeNot availableAll modelsGPT 5.4
Plugin supportYesYesNo

​
CCR + OpenRouter

This is the default provider for all new workspaces. No setup required.
CCR (Cloud Code Runtime) is Whim’s managed runtime that connects to models through OpenRouter, a multi-model API gateway. It gives you access to the widest range of models from multiple vendors — all without managing API keys. Best for:
  • Trying different models to find the best fit for your task
  • Using non-Claude/non-GPT models (Gemini, Grok, DeepSeek, etc.)
  • Getting started quickly with zero configuration
How it works:
  1. Whim manages the OpenRouter connection and API usage on your behalf
  2. API token costs are tracked as CU alongside container runtime
  3. You pick any available model when creating a task or set a workspace default
Available models include:
  • Claude Opus 4.6, Sonnet 4.5, Sonnet 4.5 (1M context), Haiku 4.5
  • GPT 5.3, GPT 5.3 Codex, GPT 5.2, GPT 5.2 Codex, GPT OSS 120B
  • Gemini 3 Pro, Gemini 2.5 Flash
  • Grok Code Fast, DeepSeek V3.2, MiniMax M2.5, Qwen3 Coder Next, Kimi K2.5
See Model Selection for the complete list.

​
Claude Subscription

Use your existing Anthropic Claude subscription to run tasks with the native Claude Code CLI. Because the Claude CLI handles API billing directly through your subscription, Whim only charges CU for container runtime — not for tokens. Best for:
  • Heavy usage where your Anthropic subscription is more cost-effective
  • Users who want the native Claude Code experience
  • Tasks that benefit from Claude Opus 4.6’s maximum capability
How it works:
  1. Connect your Claude subscription by pasting an OAuth token (see Provider Configuration)
  2. Tasks run the official Claude Code CLI natively inside the container
  3. Token costs go to your Anthropic subscription; Whim only charges CU for container time
Available models:
  • Claude Opus 4.6 (default)
  • Claude Sonnet 4.5
  • Claude Sonnet 4.5 (1M context)
  • Claude Haiku 4.5

​
Codex Subscription

Use your ChatGPT account to run tasks with OpenAI’s native Codex CLI. Like the Claude provider, token costs go through your OpenAI subscription — Whim only charges CU for container time. Best for:
  • Users with an existing ChatGPT/OpenAI subscription
  • Tasks that benefit from GPT 5.4’s capabilities
  • Users who prefer OpenAI’s reasoning and code generation style
How it works:
  1. Connect your ChatGPT account via a device code login flow (see Provider Configuration)
  2. Tasks run the official Codex CLI natively inside the container
  3. Token costs go to your OpenAI subscription; Whim only charges CU for container time
Available models:
  • GPT 5.4 (default)
  • GPT 5.3 Codex
  • GPT 5.3 Codex Spark (Preview)
  • GPT 5.2 Codex
  • GPT 5.1 Codex Mini

​
Choosing a Provider

Use CCR + OpenRouter. It’s the default, requires no configuration, and gives you access to every model.
Use Claude Subscription or Codex Subscription. With these providers, you only pay CU for container runtime. API/token costs go through your existing subscription, which can be significantly cheaper at high volume.
Use CCR + OpenRouter. It’s the only provider that gives you access to models from Anthropic, OpenAI, Google, xAI, DeepSeek, and others — all from a single provider.
Use Claude Subscription with Claude Opus 4.6, or Codex Subscription with GPT 5.4. These flagship models are available through their native CLIs for the best experience.

​
Next Steps

Model Selection

Compare models and choose the right one for your task.

Provider Configuration

Set up your provider credentials and defaults.