What is the cheapest way to run GPT-5.6 Luna?

Amazon Bedrock currently has the lowest tracked price for GPT-5.6 Luna: $0.22 input / $1.32 output per million tokens.

How many providers host GPT-5.6 Luna?

LLMap currently tracks 4 providers hosting GPT-5.6 Luna, across 7 live offers.

Provider guide

Where to run GPT-5.6 Luna

7 live offers tracked — output prices vary 4.5× between the cheapest and most expensive host, so the provider choice matters as much as the model choice.

CHEAPEST

Amazon Bedrock

$0.22 / $1.32

per 1M tokens in / out

FASTEST MEASURED

Microsoft Azure AI

125 tok/s

measured throughput · $1.10 input / $6.60 output per million tokens

ProviderIn / 1MOut / 1MThroughputContextQuantSynced

Amazon BedrockCHEAPEST$0.22 in/1M$1.32 out/1M74 tok/s1.1M—58m ago OpenRouter$0.50 in/1M$3.00 out/1M—1.1M—4d ago OpenAI$1.00 in/1M$6.00 out/1M—1.1M—5d ago OpenAIflex tier$0.25 in/1M$1.50 out/1M44 tok/s1.1M—5d ago Microsoft Azure AIzero-retention$1.00 in/1M$6.00 out/1M56 tok/s1.1M—58m ago Microsoft Azure AIzero-retention$1.10 in/1M$6.60 out/1M125 tok/s1.1M—58m ago OpenAIpriority tier$2.00 in/1M$12.00 out/1M—1.1M—8d ago

Full specs, hardware verdicts and benchmarks on theGPT-5.6 Luna model page →