By the toolist.ai editors. Researched from public sources, not hands-on tested by toolist.ai. Updated October 27, 2026; prices checked October 21, 2026.
GPT-6 Luna vs Claude Haiku 5.5
Verdict
Claude Haiku 5.5 is the stronger model at the same price, and GPT-6 Luna is the cheaper one for long prompts. Both cost $0.10 per million input tokens and $0.50 per million output tokens, and Haiku scores higher on all five Artificial Analysis scores both have, from the Intelligence Index at max effort (36 against 28) to Terminal-Bench 4.0 (33% against 13%) and Terminal-Bench-Science (20% against 9%). Luna's price rises only above 272,000 prompt tokens, to twice the input rate, while Haiku's rises fivefold above 100,000.
Verdict based on published prices, limits and benchmarks, not on our own tests.
Choose GPT-6 Luna if
- Your prompts run between 100,000 and 272,000 tokens: Luna still charges $0.10 in and $0.50 out there, while Haiku charges $0.50 and $2.50
- Your prompts pass 272,000 tokens: Luna charges $0.20 in and $0.75 out, still well under Haiku's $0.50 and $2.50
- You want typed answers from OpenAI's Decisions API, which launched in beta with GPT-6 Luna
Choose Claude Haiku 5.5 if
- You want the stronger small model on most tests: 36 against 28 on the Intelligence Index at max effort, and 25 against 18 at the default medium effort
- Your agents run terminal commands: 33% against 13% on Terminal-Bench 4.0
- You want the more recent knowledge cutoff (June 2026 against May 18, 2026)
Price
Input costs the same on both. Output costs the same on both.
| Per million tokens | GPT-6 Luna | Claude Haiku 5.5 |
|---|---|---|
| Input | $0.10 source | $0.10 source |
| Cached input | $0.01 source | $0.01 source |
| Output | $0.50 source | $0.50 source |
GPT-6 Luna: Prompts over 272,000 input tokens are billed at twice the input and cache rates and 1.5 times the output rate for the whole request. Cache writes cost $0.125 per million tokens (30-minute cache) or $0.25 (12-hour cache). Batch and Flex cost half; Fast mode costs double. OpenAI lists no Ultrafast price for it.
Claude Haiku 5.5: Prompts over 100,000 tokens are billed at $0.50 in, $2.50 out and $0.05 for cache reads per million tokens for the whole request; the prompt length counts cache reads and writes. Cache writes cost $0.125 per million tokens (5-minute cache) or $0.20 (1-hour cache), or $0.625 and $1 for prompts over 100,000 tokens. The Batch API costs half.
Worked example: what a month costs
We priced three monthly workloads at the standard rates. The provider does not publish these figures.
| Workload | GPT-6 Luna | Claude Haiku 5.5 |
|---|---|---|
| Support chatbot 100,000 replies, each 3,000 input and 400 output tokens, no caching (300M in, 40M out) | $50.00 | $50.00 |
| Document Q&A 5,000 questions over the same 60,000-token document set (90% cache hits), 800 output tokens each (300M in, 4M out) | $7.70 | $7.70 |
| Coding agent 200 tasks, each 1.5 million input tokens (90% cache hits) and 50,000 output tokens (300M in, 10M out) | $10.70 | $10.70 |
Price your own workload
| Model | Cost a month |
|---|---|
| GPT-6 Luna | $7.50 |
| Claude Haiku 5.5 | $7.50 |
Standard API prices, no batch or priority discounts. Cached input is billed at the cache-read price; cache-write charges are not included. Reasoning models bill their thinking as output tokens.
Limits and release dates
| GPT-6 Luna | Claude Haiku 5.5 | |
|---|---|---|
| Released | September 22, 2026 source | October 7, 2026 source |
| Context window | 1,050,000 tokens source | 1,000,000 tokens source |
| Maximum output | 128,000 tokens source | 128,000 tokens source |
| Knowledge cutoff | May 18, 2026 source | June 2026 source |
| Input and output | Text and images in; text out source | Text and images in; text out source |
| API model name | gpt-6-luna | claude-haiku-5-5 |
Benchmarks
Scores as published by the provider or measured by Artificial Analysis; higher is better. A dash means no score was published.
| Benchmark | GPT-6 Luna | Claude Haiku 5.5 |
|---|---|---|
| Artificial Analysis Intelligence Index max effort | 28.0 source | 36.0 source |
| Intelligence Index at everyday effort medium effort, the API default | 18.0 source | 25.0 source |
| Terminal-Bench 4.0 max effort | 13.0% source | 33.0% source |
| HLE-Diamond max effort | 22.0% source | 30.0% source |
| Terminal-Bench-Science 0.1 max effort | 9.0% source | 20.0% source |
| Output speed | 124.3 tokens/s (max effort, Artificial Analysis; no medium-effort figure is published) source | 175.1 tokens/s (medium effort, Artificial Analysis) source |
Benchmark and speed data from Artificial Analysis where linked.
Read more
- GPT-6 Luna: price, context window and benchmarks
- Claude Haiku 5.5: price, context window and benchmarks
- Claude Haiku 5.5 vs Claude Sonnet 5.5
- GPT-6.1 Sol vs GPT-6 Luna
- AI model API price table
Sources
- GPT-6.1 Sol vs GPT-6 Luna (checked 2026-10-27)
- OpenAI API docs: GPT-6 Luna (checked 2026-10-21)
- OpenAI API docs: Pricing (checked 2026-10-21)
- OpenAI API docs: Changelog (checked 2026-10-21)
- OpenAI API docs: Using GPT-6 (checked 2026-10-21)
- ChatGPT docs: Models (checked 2026-10-21)
- ChatGPT docs: Pricing and plan access (checked 2026-10-21)
- GPT-6 Luna (max) (checked 2026-10-27)
- GPT-6 Luna (medium) (checked 2026-10-27)
- OpenRouter: GPT-6 Luna (checked 2026-10-21)
- Claude Haiku 5.5 vs Claude Sonnet 5.5 (checked 2026-10-27)
- Anthropic: Introducing Claude Haiku 5.5 (checked 2026-10-21)
- Claude docs: Claude Haiku 5.5 (checked 2026-10-21)
- Claude docs: What's new in Claude Haiku 5.5 (checked 2026-10-21)
- Claude docs: Effort (checked 2026-10-21)
- Claude docs: Pricing (checked 2026-10-21)
- Claude plans and pricing (checked 2026-10-21)
- Claude docs: API and data retention (checked 2026-10-21)
- Claude Haiku 5.5 (max) (checked 2026-10-27)
- Claude Haiku 5.5 (medium) (checked 2026-10-27)
- OpenRouter: Claude Haiku 5.5 (checked 2026-10-21)