Published rates rank differently once prompt caching, batch discounts and your own input-to-output ratio are applied. Run both against your numbers.
Claude Haiku 5.5 vs GPT-6 Luna
Identical $0.10 / $0.50 list prices at the cheap tier.
Once Haiku 5.5 prompts pass 100K tokens, they bill at 5x.
Monthly cost by workload shape
The same volume costs very different amounts depending on the shape of the traffic. No caching or batch applied.
| Workload | Claude Haiku 5.5 | GPT-6 Luna | Difference | |
|---|---|---|---|---|
Customer support chatbot 50K/day · 800 in / 300 out | $350.06 | $350.06 | Same cost | Model it → |
RAG / search answers 20K/day · 3000 in / 400 out | $304.40 | $304.40 | Same cost | Model it → |
Agent / tool-use loop 8.0K/day · 2500 in / 600 out | $133.94 | $133.94 | Same cost | Model it → |
Batch summarization 100K/day · 5000 in / 500 out | $2,283.00 | $2,283.00 | Same cost | Model it → |
Code assistant 15K/day · 2000 in / 800 out | $273.96 | $273.96 | Same cost | Model it → |
Frequently asked questions
Which is cheaper, Claude Haiku 5.5 or GPT-6 Luna?
Neither, on list price: both charge $0.100 input and $0.500 output per 1M tokens, so they cost the same at every input:output ratio. Caching, batch and long-context terms decide the bill.
What is the price difference between Claude Haiku 5.5 and GPT-6 Luna?
Claude Haiku 5.5: $0.100 input, $0.500 output per 1M tokens. GPT-6 Luna: $0.100 input, $0.500 output. That is 1.0x on input and 1.0x on output.
Does prompt caching change the answer?
Claude Haiku 5.5 caches input at $0.010/1M and GPT-6 Luna caches at $0.010/1M. For input-heavy workloads with repeated context, caching can matter more than the base rate difference, model both in the calculator rather than comparing rate cards.
This page compares published prices only. It makes no claim about which model performs better on any task; that depends entirely on your evaluations.
Rate cards decide nothing on their own.
Most teams run several models at once. Finitizer Tokenomics shows what each one actually costs you, read from your own bills and exports, which workloads could test a smaller model, what they would cost on an open-weight model, and, for models you serve yourself, the cost per token beside the API price. A routing decision starts from your spend rather than a price page.
See how Finitizer works