BedrockExplorer

Claude Haiku 4.5 vs GPT-5.6 Luna on Amazon Bedrock

In us-east-1, GPT-5.6 Luna costs $0.200 per 1M input tokens against $1.00 — a 5.0× difference. Prices, context, capabilities and region coverage below.

Claude Haiku 4.5 Anthropic
anthropic.claude-haiku-4-5-20251001-v1:0
Input / 1M
$1.00
Output / 1M
$5.00
Context
200K
Est. / month
$1,170
GPT-5.6 Luna OpenAI
openai.gpt-5.6-luna
Input / 1M
$0.200
Output / 1M
$1.20
Context
1M
Est. / month
$252.00

Capabilities

Capability Claude Haiku 4.5 GPT-5.6 Luna
Tool use ✓ ✓
Vision ✓ ✓
Prompt caching ✓ ✓
Batch inference ✓ —
Streaming ✓ ✓

Region availability

Region Claude Haiku 4.5 GPT-5.6 Luna
us-east-1 US East (N. Virginia) Inference profile only Inference profile only
us-east-2 US East (Ohio) Inference profile only Inference profile only
us-west-1 US West (N. California) Inference profile only Inference profile only
us-west-2 US West (Oregon) Inference profile only Inference profile only
ca-central-1 Canada (Central) Inference profile only Inference profile only
ca-west-1 Canada West (Calgary) Inference profile only Inference profile only
mx-central-1 Mexico (Central) Inference profile only Not available
eu-west-1 EU (Ireland) Inference profile only Inference profile only
eu-west-2 EU (London) Inference profile only Inference profile only
eu-west-3 EU (Paris) Inference profile only Inference profile only
eu-central-1 EU (Frankfurt) Inference profile only Inference profile only
eu-central-2 Europe (Zurich) Inference profile only Inference profile only
eu-north-1 EU (Stockholm) Inference profile only Inference profile only
eu-south-1 EU (Milan) Inference profile only Inference profile only
eu-south-2 Europe (Spain) Inference profile only Inference profile only
ap-east-2 Asia Pacific (Taipei) Inference profile only Inference profile only
ap-northeast-1 Asia Pacific (Tokyo) Inference profile only Inference profile only
ap-northeast-2 Asia Pacific (Seoul) Inference profile only Inference profile only
ap-northeast-3 Asia Pacific (Osaka) Inference profile only Inference profile only
ap-south-1 Asia Pacific (Mumbai) Inference profile only Inference profile only
ap-south-2 Asia Pacific (Hyderabad) Inference profile only Inference profile only
ap-southeast-1 Asia Pacific (Singapore) Inference profile only Inference profile only
ap-southeast-2 Asia Pacific (Sydney) Inference profile only Inference profile only
ap-southeast-3 Asia Pacific (Jakarta) Inference profile only Inference profile only
ap-southeast-4 Asia Pacific (Melbourne) Inference profile only Inference profile only
ap-southeast-5 Asia Pacific (Malaysia) Inference profile only Inference profile only
ap-southeast-6 Asia Pacific (New Zealand) Inference profile only Not available
ap-southeast-7 Asia Pacific (Thailand) Inference profile only Inference profile only
il-central-1 Israel (Tel Aviv) Inference profile only Inference profile only
af-south-1 Africa (Cape Town) Inference profile only Not available
sa-east-1 South America (Sao Paulo) Inference profile only Inference profile only

28 of 31 regions support both models.

Common questions

Is Claude Haiku 4.5 or GPT-5.6 Luna cheaper on Amazon Bedrock?
GPT-5.6 Luna is cheaper on input tokens: $0.200 against $1.00 per 1M in us-east-1, a 5.0× difference. At 2,000 runs a day with 12,000 input and 1,500 output tokens each, that is $252.00 against $1,170 a month before any caching or batch discount.
Does Claude Haiku 4.5 or GPT-5.6 Luna have the larger context window?
GPT-5.6 Luna, at 1M tokens against 200K.
Which regions support Claude Haiku 4.5 and GPT-5.6 Luna?
Claude Haiku 4.5 runs in 31 of the 31 AWS regions tracked here and GPT-5.6 Luna in 28. 28 regions support both, so a workload that may switch between them should target one of those.
Model your own volume → Open in interactive compare →

People comparing these also look at

Other pairs involving Claude Haiku 4.5 or GPT-5.6 Luna.