Model comparison
Claude Haiku 5.5 vs GPT-6 Luna: pricing, context and features compared
Claude Haiku 5.5 costs $0.10 per million input tokens and $0.50 per million output tokens; GPT-6 Luna costs $0.10 and $0.50. Here is how they compare on specs, on what real workloads cost, and on which to pick for what, from prices updated daily.
Haiku and Luna are Anthropic’s and OpenAI’s small, fast models, used for routing, extraction, summaries and sub-agents, where every request is cheap but there are a lot of them.
- Input / 1M
- $0.10
- Output / 1M
- $0.50
- Context
- 1,000,000
- Max output
- 128,000
- Input / 1M
- $0.10
- Output / 1M
- $0.50
- Context
- 1,050,000
- Max output
- 128,000
Verdict
Claude Haiku 5.5 or GPT-6 Luna: which to pick
Claude Haiku 5.5 and GPT-6 Luna have about the same list prices per token. On our data, GPT-6 Luna has the edge for very long prompts, and Claude Haiku 5.5 has no measurable edge; its case rests on answer quality, so choose it only if it does clearly better in your own tests. For scale, a support chatbot handling 1,000 requests a day costs about $8.59 a month on Claude Haiku 5.5 and $8.59 on GPT-6 Luna.
- Lowest bill for typical workloadsEither
- Both cost about the same on all 4 workloads below.
- Very long prompts (whole codebases, long documents)Pick GPT-6 Luna
- Both read about 1,000,000 tokens per request. One 800,000-token prompt with a 2,000-token reply costs $0.41 on Claude Haiku 5.5 and $0.16 on GPT-6 Luna. Claude Haiku 5.5 charges $0.50 / $2.50 per 1M once a prompt passes 100,000 tokens. GPT-6 Luna charges $0.20 / $0.75 per 1M once a prompt passes 272,000 tokens.
- Long replies (reports, large code files)Either
- Both cap a single reply at about 128,000 tokens.
- Images, audio, video or files in the promptEither
- Both accept images and files as well as text.
- Repeated long prompts (prompt caching)Either
- Both bill cached input at about $0.01 per 1M tokens.
- Overnight batch jobsEither
- Both offer batch prices and cost about the same on the document summariser workload.
- Self-hosting or choosing your own hostEither
- Neither has downloadable weights: both are only available through their provider’s API and cloud partners.
Criteria use only list prices, published limits and the features each provider declares. We don’t rank answer quality; test both models on your own prompts before you commit.
Specs
Claude Haiku 5.5 vs GPT-6 Luna side by side
| Spec | Claude Haiku 5.5 | GPT-6 Luna |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Input / 1M tokens | $0.10 | $0.10 |
| Output / 1M tokens | $0.50 | $0.50 |
| Cached input / 1M | $0.01 | $0.01 |
| Cache write / 1M | $0.125 | $0.125 |
| Batch in / out | $0.05 / $0.25 | $0.05 / $0.25 |
| Long-prompt price | $0.50 / $2.50 over 100,000 | $0.20 / $0.75 over 272,000 |
| Context window | 1,000,000 | 1,050,000 (better) |
| Max output | 128,000 | 128,000 |
| Accepts | text, images, files | text, images, files |
| Image input | Yes | Yes |
| Tool calling | Yes | Yes |
| Reasoning mode | Yes | Yes |
| Prompt caching | Yes | Yes |
| Structured output | Yes | Yes |
| Open weights | No | No |
| Knowledge cutoff | Not published | Not published |
| First listed | 2026-10-07 | 2026-09-22 |
Prices in US dollars per million tokens. Highlighted: the lower price or larger limit where the difference is over 5%. Features are as declared by each provider’s API; “First listed” is when our data source first listed the model. Long-prompt prices apply to the whole request once the prompt passes the threshold.
Costs
What Claude Haiku 5.5 and GPT-6 Luna cost for real workloads
| Workload | Tokens in / out | Requests/day | Claude Haiku 5.5 / month | GPT-6 Luna / month | Cheaper |
|---|---|---|---|---|---|
| Support chatbotCalculator: Claude Haiku 5.5 · GPT-6 Luna | 1,500 / 400 | 1,000 | $8.5950% cached | $8.5950% cached | About the same |
| RAG appCalculator: Claude Haiku 5.5 · GPT-6 Luna | 6,000 / 500 | 500 | $11.2820% cached | $11.2820% cached | About the same |
| Coding agentCalculator: Claude Haiku 5.5 · GPT-6 Luna | 40,000 / 2,000 | 200 | $12.9080% cached | $12.9080% cached | About the same |
| Document summariserCalculator: Claude Haiku 5.5 · GPT-6 Luna | 8,000 / 600 | 300 | $5.02batch | $5.02batch | About the same |
Each row uses the same token counts for both models and list prices, with caching and batch discounts only where the provider publishes them. A month is 365 ÷ 12 days. The support chatbot reads 50% of its prompt from the cache. The RAG app reads 20% of its prompt from the cache. The coding agent reads 80% of its prompt from the cache. The document summariser runs as a batch job where a batch price exists. Open either model in the LLM cost calculator to change any number.
Tokenizers
Are their per-token prices comparable?
Not exactly. Each model splits text into tokens its own way, so the same prompt can be a different number of tokens on Claude Haiku 5.5 and GPT-6 Luna, and the cheaper per-token price isn’t always the cheaper bill. Where a tokenizer isn’t public or measured we say so rather than guess.
| Model | Tokenizer | Tokens per 1,000 words | Output per 1M words |
|---|---|---|---|
| Claude Haiku 5.5 | Claude (Opus 4.7 and later)Provider’s published figure | 1,802 | $0.90 |
| GPT-6 Luna | OpenAI hasn’t published the tokenizer GPT-6 Luna uses, so its count for a given text isn’t known exactly. | ||
English prose, measured on the Universal Declaration of Human Rights (2026-10-11) or taken from the provider’s published words-per-token figure. Code, JSON and other languages use more tokens per word. Count your own text in the token counter or convert with tokens to words.
FAQ
Claude Haiku 5.5 vs GPT-6 Luna questions
Is Claude Haiku 5.5 cheaper than GPT-6 Luna?
Claude Haiku 5.5 costs $0.10 per million input tokens and $0.50 per million output tokens; GPT-6 Luna costs $0.10 and $0.50. For a support chatbot handling 1,000 requests a day, that is about $8.59 a month on Claude Haiku 5.5 against $8.59 on GPT-6 Luna, so neither is meaningfully cheaper there. Try your own numbers in the LLM cost calculator.
Which has the bigger context window, Claude Haiku 5.5 or GPT-6 Luna?
GPT-6 Luna has the slightly larger context window. Claude Haiku 5.5 reads up to 1,000,000 tokens (roughly 750,000 English words) and writes up to 128,000 tokens per reply. GPT-6 Luna reads up to 1,050,000 tokens (roughly 787,500 English words) and writes up to 128,000 tokens per reply. The window is shared between your prompt and the reply. Word counts use a rough 0.75 words per token; real ratios depend on the tokenizer.
Is Claude Haiku 5.5 or GPT-6 Luna better for coding?
We don’t publish benchmark scores, so this page can’t say which writes better code. What we can show is cost: a coding agent re-sending 40,000 tokens of context (80% cached) and writing 2,000 tokens, 200 times a day, costs about $12.90 a month on Claude Haiku 5.5 and $12.90 on GPT-6 Luna. Both support tool calling. Test both on tasks from your own repository.
Can Claude Haiku 5.5 and GPT-6 Luna read images?
Claude Haiku 5.5 accepts images and files as well as text. GPT-6 Luna accepts images and files as well as text. So both can read screenshots, photos and scanned pages. This is what each provider declares for its API, not a measure of how well it works.
Are Claude Haiku 5.5 and GPT-6 Luna open source?
No. Neither Claude Haiku 5.5 nor GPT-6 Luna has published weights: you can only use them through Anthropic’s and OpenAI’s APIs and their cloud partners. If you need a model you can host yourself, compare open-weight options in the model comparison table.
Do Claude Haiku 5.5 and GPT-6 Luna count tokens the same way?
Not necessarily. Each model splits text into tokens with its own tokenizer, so the same prompt can be a different number of tokens on each, and per-token prices aren’t directly comparable. Claude Haiku 5.5 uses about 1,802 tokens per 1,000 English words (the provider’s published figure). OpenAI hasn’t published the tokenizer GPT-6 Luna uses, so its count for a given text isn’t known exactly. Count your own text with the token counter or convert with tokens to words.