Pricing

Access Tiers

One prompt. Every model.
Choose how far to push it.

Fan out a single question to multiple LLMs simultaneously and compare responses. Your token budget determines how much you can explore each period.

Token Currency 1,000,000
All Plans Include Judge LLM analysis
Enforcement Server-side · Exact counts
Providers OpenAI · Anthropic · Gemini · Grok
Active
Red Cup

Anonymous. No sign-up.

Free

no account

Token Budget 1,000,000
Lanes 3 max
Max output 512 tok
Premium models ✗ locked
Custom prompt ✗ no
Share sessions ✗ no
Reqs / hour 20
Reset cadence session-scoped
Active
Open Bar

Free forever with a WP account.

Free

verified email

Token Budget 1,000,000
Lanes 3 max
Max output 1,024 tok
Premium models ✗ locked
Custom prompt ✓ yes
Share sessions ✓ yes
Reqs / hour 60
Reset cadence monthly
Active
Run A Tab

Fresh tokens every month. Cancel anytime.

$3 /mo

/month

Token Budget 1,000,000
Lanes 8 max
Max output 4,096 tok
Premium models ✓ unlocked
Custom prompt ✓ yes
Share sessions ✓ yes
Reqs / hour 200
Reset cadence billing date
Full Specification Matrix
Parameter 🥤 Red Cup 🍺 Open Bar 💵 Cash Bar 🪙 Run A Tab
Identity & Cost
Auth requirement none WP + email WP + purchase WP + subscription
Price free free $5 one-time $3 /month
Token Budget
Allotment 1M 1M 1M 1M
Period session month until spent billing cycle
Reset cadence session-scoped monthly never expires billing date
Comparison Engine
Max lanes 3 3 8 8
Max output tokens 512 1,024 4,096 4,096
Premium models
Custom system prompt
Share sessions
Judge LLM
Rate Limits (per hour)
Requests 20 60 200 200
Comparisons 5 20 100 100
Frequently Asked
What counts as a token?

Roughly 4 characters or ¾ of a word. Every prompt you send and every response you receive consumes tokens. The judge LLM analysis also draws from your budget. One million tokens is substantial — equivalent to ~750k words processed.

What are "premium models"?

Currently claude-opus-4-6 and gpt-4-turbo — the most capable model from each provider. Unlocked for Cash Bar and Run A Tab. All other models (GPT-4o-mini, Claude Haiku, Gemini, Grok) are available on every tier.

What happens when tokens run out?

New comparisons are blocked until balance is replenished. Free tiers reset on a schedule; Cash Bar tokens never expire — purchase another pack when empty. Run A Tab users get fresh tokens automatically on each billing date.

Can I stack Cash Bar packs with a subscription?

Yes. If you have both run_a_tab and cash_bar balance, subscription tokens are consumed first — your Cash Bar credits are preserved until the subscription lapses. Cash Bar can be purchased multiple times.

How do I cancel a subscription?

Manage billing through the Stripe Customer Portal — accessible from your account settings. Cancel, update payment method, or download invoices there. On cancellation, tier reverts to Open Bar at period end.

Is my data used to train models?

VoxDiff proxies requests to each provider under API terms. Your prompts are subject to each provider's data policy — we do not feed your sessions to any training pipeline of our own. Check OpenAI and Anthropic API policies for provider-level detail.