Claude 3 Haiku
Fast, affordable model with strong vision capabilities and performance for diverse enterprise applications.
Anthropic 
Amazon Bedrock Fast, affordable text and vision processing
Claude 3 Haiku is a text generation model developed by Anthropic, positioned as the fastest and most affordable model in the Claude 3 family. It features a 200,000-token context window and vision capabilities, making it suitable for tasks that require processing large documents or analyzing images alongside text. The model's training data has a cutoff of August 2023.
Haiku is designed for enterprise use cases where throughput and cost efficiency matter, such as customer support, real-time chat, and batch processing of large datasets. It is capable of processing approximately 21,000 tokens — roughly 30 pages — per second for prompts under 32,000 tokens, which makes it well-suited for latency-sensitive applications and workloads that involve running many smaller tasks in parallel.
What Claude 3 Haiku supports
Large Context Window
Accepts prompts up to 200,000 tokens, enabling analysis of long documents, codebases, or multi-turn conversations in a single request.
Vision Understanding
Processes images alongside text input, supporting tasks like document analysis, chart interpretation, and visual question answering.
High-Speed Inference
Processes approximately 21,000 tokens per second for prompts under 32,000 tokens, enabling low-latency responses for real-time applications.
Text Generation
Generates coherent, contextually relevant text for tasks including summarization, classification, drafting, and question answering.
Instruction Following
Responds to structured prompts and multi-step instructions, supporting task automation and enterprise workflow integration.
Ready to build with Claude 3 Haiku?
Get Started FreeBenchmark scores
Scores represent accuracy — the percentage of questions answered correctly on each test.
| Benchmark | What it tests | Score |
|---|---|---|
| GPQA Diamond | PhD-level science questions (biology, physics, chemistry) | 37.4% |
| MATH-500 | Undergraduate and competition-level math problems | 39.4% |
| AIME 2024 | American math olympiad problems | 1.0% |
| LiveCodeBench | Real-world coding tasks from recent competitions | 15.4% |
| HLE | Questions that challenge frontier models across many domains | 3.9% |
| SciCode | Scientific research coding and numerical methods | 18.6% |
Common questions about Claude 3 Haiku
What is the context window size for Claude 3 Haiku?
Claude 3 Haiku supports a context window of 200,000 tokens, which is approximately equivalent to several hundred pages of text in a single prompt.
What is the training data cutoff for Claude 3 Haiku?
The model's training data has a cutoff of August 2023, meaning it does not have knowledge of events or information published after that date.
Does Claude 3 Haiku support image inputs?
Yes, Claude 3 Haiku includes vision capabilities, allowing it to process and reason about images submitted alongside text prompts.
What types of tasks is Claude 3 Haiku best suited for?
Haiku is designed for latency-sensitive and cost-sensitive enterprise applications such as customer support, real-time chat, document processing, and running many small tasks in parallel.
Who publishes Claude 3 Haiku?
Claude 3 Haiku is published by Anthropic, an AI safety company. It is part of the Claude 3 model family, which also includes Claude 3 Sonnet and Claude 3 Opus.
What people think about Claude 3 Haiku
The available Reddit thread focuses on Claude Opus 4.5 rather than Claude 3 Haiku specifically, so direct community sentiment about Haiku is limited in this dataset. General community discussion around the Claude 3 family tends to highlight Haiku's speed and cost efficiency as practical advantages for high-volume tasks.
Developers commonly reference Haiku for use cases where response latency and API cost are primary constraints, such as chatbots and automated pipelines. Concerns in broader Claude 3 discussions often center on capability trade-offs relative to larger models in the same family.
New benchmark: Claude Opus 4.5 broke the efficiency wall.+21% intelligence while getting 66% cheaper
Documentation & links
Parameters & options
Explore similar models
Anthropic Claude 4.6 Sonnet
Anthropic's most capable Sonnet model yet, delivering near-Opus-level intelligence for coding, agents, and computer use with a 1M token context window.
Anthropic Claude 4.6 Opus
Anthropic's most intelligent model, setting new standards in agentic coding, complex reasoning, and professional knowledge work.
Anthropic Claude 4.5 Opus
Anthropic's most powerful flagship model, setting the state of the art in coding, agentic workflows, and real-world computer use.
Google Gemini 3.1 Pro
Google's frontier reasoning model delivering enhanced software engineering, agentic reliability, and multimodal intelligence across a 1M-token context window.
OpenAI GPT‑5.2 Pro
The most capable model series yet for professional knowledge work.
Nvidia Nemotron 3 Nano 30B
NVIDIA's compact open-weight hybrid MoE model with 30B total / 3.5B active parameters, delivering strong reasoning and coding performance up to 1M context.
Start building with Claude 3 Haiku
No API keys required. Create AI-powered workflows with Claude 3 Haiku in minutes — free.