LLM model

Claude Haiku

Fast small Claude for latency-sensitive tasks.

Reference Cloud Wizard AWS
Open the selection wizard All models
Illustrative specs. Context window, modalities, output cost, and EU availability for Claude Haiku are representative, last verified 1 Jun 2026. Verify against the provider before committing.

Claude Haiku at a glance

The fast/cheap tier of the Claude family on Bedrock.

Hosting Bedrock Bedrock (AWS)
Context window 200,000 tokens
Modalities text, vision inputs
Output cost $4 / 1M tokens

Best for

  • Realtime assistants
  • High-volume classification

Is it the right model?

Match it against your requirements.

The selection wizard ranks Claude Haiku against every other model for your latency, context, modality, cost, and residency needs — and shows where it wins and where something else fits better.

Open the LLM Selection Wizard

Frequently asked questions

Claude Haiku specs and cost.

What is Claude Haiku best for? Fast small Claude for latency-sensitive tasks. It fits Realtime assistants, High-volume classification.
What context window and modalities does Claude Haiku support? Claude Haiku handles up to 200,000 tokens of context and supports text, vision input. It runs on Bedrock (AWS).
How much does Claude Haiku cost? Around $4 per 1M output tokens (illustrative, verified 1 Jun 2026). Output tokens usually dominate the bill — verify input and cached pricing against the provider before budgeting.