Claude Haiku 5.5 vs GPT-6 Luna
A side-by-side of live API prices, providers, speed, usage and benchmarks. No verdict: the right choice depends on your workload.
Claude Haiku 5.5 costs $0.10 per 1M input and $0.50 per 1M output tokens at the standard tier, via Anthropic; GPT-6 Luna costs $0.10 and $0.50, via OpenAI. At a 3:1 input:output mix, Claude Haiku 5.5 is the same price as GPT-6 Luna.
Why these two are compared
The small, fast tiers of Anthropic and OpenAI: Claude Haiku 5.5 for high-volume, cost-sensitive work such as summarisation and subagents, GPT-6 Luna for high-volume and latency-sensitive workloads such as chat and classification.
Side by side
| Measure | Claude Haiku 5.5 | GPT-6 Luna |
|---|---|---|
| Developer | Anthropic | OpenAI |
| Released | 2026-10-07 | 2026-09-22 |
| Standard price, input / output per 1M | $0.10 / $0.50 | $0.10 / $0.50 |
| Blended price (3:1) | $0.20 | $0.20 |
| Cheapest standard provider | Anthropic | OpenAI |
| Cache read per 1M | $0.01 | $0.01 |
| Providers at the standard tier | 5 | 3 |
| Context window (tokens) | 1,000,000 | 1,050,000 |
| Max output (tokens) | 128,000 | 128,000 |
| Input modalities | text, image, file | file, image, text |
| Long-prompt pricing | Prompts over 100,000 tokens: $0.50 input / $2.50 output per 1M tokens | Prompts over 272,000 tokens: $0.20 input / $0.75 output per 1M tokens |
| Fastest endpoint | 173 tok/sGoogle (Global endpoint) | 106 tok/sOpenAI (Fast tier) |
| Lowest latency | 354 msGoogle (Europe region) | 849 msAzure (EU region) |
| OpenRouter tokens (2026-10-16) | 302.61B | 988.36B |
| AA intelligence | 43.4 | 38.1 |
| AA coding | Not reported | Not reported |
| AA agentic | Not reported | Not reported |
“Not reported” means the source publishes no value. AA scores are the Artificial Analysis indexes reported through OpenRouter (CC BY 4.0), not SF·SIX measurements.
Cost for three example workloads
| Workload | Claude Haiku 5.5 / month | GPT-6 Luna / month |
|---|---|---|
| Support chatbot50,000 requests × 1,500 in / 400 out tokens | $17.50 | $17.50 |
| RAG / research assistant20,000 requests × 8,000 in / 700 out tokens | $23.00 | $23.00 |
| Batch summarisation10,000 requests × 10,000 in / 500 out tokens | $12.50 | $12.50 |
Standard headline prices; caching and long-context surcharges not included. Price your own workload.
Claude Haiku 5.5 providers
| Provider and terms | Input / 1M | Output / 1M | Throughput | Latency |
|---|---|---|---|---|
| Amazon Bedrock | $0.10 | $0.50 | 111 tok/s | 2.72 s |
| Anthropic | $0.10 | $0.50 | 128 tok/s | 718 ms |
| AzureGlobal endpoint | $0.10 | $0.50 | 137 tok/s | 1.60 s |
| Claude Platform on AWS | $0.10 | $0.50 | 154 tok/s | 747 ms |
| GoogleGlobal endpoint | $0.10 | $0.50 | 173 tok/s | 1.60 s |
| Amazon Bedrockeu-west-1 region | $0.11 | $0.55 | Not reported | Not reported |
| Amazon Bedrockus-east-1 region | $0.11 | $0.55 | 85 tok/s | 1.40 s |
| AzureUS region | $0.11 | $0.55 | Not reported | Not reported |
| GoogleEurope region | $0.11 | $0.55 | 155 tok/s | 354 ms |
| GoogleUS region | $0.11 | $0.55 | 64 tok/s | 1.23 s |
GPT-6 Luna providers
| Provider and terms | Input / 1M | Output / 1M | Throughput | Latency |
|---|---|---|---|---|
| Azure | $0.10 | $0.50 | 43 tok/s | 5.47 s |
| OpenAI | $0.10 | $0.50 | 71 tok/s | 2.63 s |
| Amazon Bedrockus-east-1 region | $0.11 | $0.55 | 84 tok/s | 950 ms |
| AzureEU region | $0.11 | $0.55 | 53 tok/s | 849 ms |
| AzureUS region | $0.11 | $0.55 | 28 tok/s | 6.32 s |
| OpenAIFast tier | $0.20 | $1.00 | 106 tok/s | 1.08 s |
| OpenAIFlex tier | $0.05 | $0.25 | 91 tok/s | 2.24 s |
Sources and method
Prices are USD per 1M tokens at the standard service tier in the default region: the cheapest qualifying provider per model, with flex, batch, priority and fast endpoints and unusually low outliers never used as the headline. “Blended” weights input and output 3:1. Prices: OpenRouter, observed at the times on each model page. Speed: OpenRouter, previous 30 minutes. Usage: OpenRouter rankings, latest full UTC day (CC BY 4.0).