Pay-per-use

Simple, transparent pricing

Credit-based billing for GPT, Claude, Gemini, and 9+ models. Start free with 20 credits. Pay only for the tokens you actually consume. No subscription required.

Free Trial
$0
20 credits · Free
One-time trial credits
Start free — no card required
Starter
$3
300 credits · $0.01/credit
Try paid — less than a coffee
Start free first
Standard
$10
1,100 credits · $0.0091/credit
10% bonus · paid credits never expire
Start free first
Power
$25
3,000 credits · $0.0083/credit
20% bonus · best value for Compare & Blend
Start free first

Credits are settled by actual token usage (input + output), selected model, and mode. Message volume is not a fixed guarantee.

Credit cost per mode

Each mode reserves a base number of credits per request. Final cost is settled by actual token usage — frontier models cost more per credit, cheaper models cost less.

ModeCreditsDescription
Chat1Single model prompt + response
Compare3Same prompt to multiple models
Blend4Gather + synthesize from multiple models
Judge5Models evaluate each other's outputs
Failover1Chat with automatic fallback chain

Cost examples

Casual developer
50 Chat requests/month with GPT-5.2 Mini
50 × 1 credit = 50 credits → covered by Starter ($3)
Startup team
200 Chat + 30 Compare requests/month
200 × 1 + 30 × 3 = 290 credits → one Starter pack ($3)
Production API
1,000 Chat + 100 Compare + 50 Blend/month
1,000 + 300 + 200 = 1,500 credits → one Power pack ($25, 3,000 cr)

LLMWise vs alternatives

FeatureLLMWiseDirect APIChatGPT Plus
Monthly minimum$0 (free trial)$0$20/mo
Model access9+ models, all providers1 provider per keyGPT models only
Failover routingBuilt-in meshBuild it yourselfNo
Billing modelCredit-based, usage-settledPer-token per-providerFlat subscription
BYOK (own keys)YesN/A (you are the provider)No
Data controlZero-retention modeVaries by providerLimited

Model-specific pricing

GPT-5.2 Pricing
Claude Pricing
Gemini Pricing
DeepSeek Pricing
Llama Pricing
Grok Pricing

Pricing FAQ

What are credits?
Credits are LLMWise's billing unit. Each API request reserves credits based on the mode (Chat=1, Compare=3, etc.), then final cost is settled by actual token usage. This means you pay for what you consume, not a flat rate.
Do credits expire?
Free trial credits are available for 7 days. Paid credits never expire — use them whenever you need them.
What is usage-settlement?
When you make a request, LLMWise reserves the base credit cost (e.g. 1 credit for Chat). After the response completes, the actual token cost is calculated and any difference is reconciled. Frontier models may cost more per credit; cheaper models cost less.
Are any models free?
Some open-source models are marked as free in LLMWise and don't deduct credits. Check the model selector in the dashboard for current free models.
What is BYOK?
Bring Your Own Keys — you can add your own OpenAI, Anthropic, or Google API keys. Requests route directly to the provider and skip credit charges entirely. You still get failover and orchestration.
Is there enterprise pricing?
For teams needing custom limits, team billing, procurement support, or SLAs, contact us at hello@llmwise.ai.
How do I add credits?
Sign in to your dashboard and visit the Credits page. Choose a credit pack, pay with any card via Stripe, and credits are added instantly. Auto top-up is also available.

Start free — 20 credits

No credit card required. Paid credits never expire. Cancel anytime.

Start free with 20 creditsSee API docs