Claude 3 Haiku
VerifiedFast multimodal model for quick text and image analysis.
About Claude 3 Haiku
Claude 3 Haiku was designed by Anthropic as the most compact member of the Claude 3 family. It combines text and image modalities in a single system while maintaining a 200,000-token context window. The model is not released as open weights.
Its architecture emphasizes speed and cost efficiency without sacrificing core reasoning capabilities. This makes it suitable for workloads that require rapid iteration over visual and textual data. Users benefit from consistent performance across varied input types.
Typical applications include real-time customer support, content moderation pipelines, and quick document or image summarization. Developers often integrate it when low latency matters more than maximum model size.
Capabilities
Benchmarks & performance
Independent evaluation scores and measured speed.
Source: Artificial Analysis
How Claude 3 Haiku compares
Claude 3 Haiku (striped bar) vs other multimodal on intelligence, speed and price.
Intelligence
Artificial Analysis Intelligence Index · Higher is better · Claude 3 Haiku ranks #84 of 88
Price
USD per 1M output tokens · Lower is better · Claude 3 Haiku ranks #47 of 155
Sources: Artificial Analysis (intelligence, speed) · OpenRouter (price).
Best for
Analyzing Long Documents
Claude 3 Haiku leverages its 200000-token context window for long-context reasoning and document summarization across extensive inputs.
Processing Multimodal Inputs
It delivers multimodal text and image understanding, enabling accurate interpretation of combined visual and textual data in a single query.
Supporting Code Tasks
The model provides fast text generation along with code assistance and instruction following for programming and development workflows.
Strengths & limitations
Strengths
- +High speed and low latency
- +Cost-effective for high-volume use
- +Strong safety alignments
- +Reliable everyday task handling
Limitations
- –Less depth on complex reasoning than larger models
- –Basic vision capabilities
- –Reduced performance on specialized or creative tasks
Cost calculator
Estimate what Claude 3 Haiku would cost for your usage.
Based on Claude 3 Haiku's $0.25/1M input · $1.25/1M output. Estimate only — actual cost varies by provider and caching.
Quick start
OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. Only the model slug changes between models.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://openrouter.ai/api/v1",
apiKey: process.env.OPENROUTER_API_KEY,
});
const completion = await client.chat.completions.create({
model: "anthropic/claude-3-haiku",
messages: [{ role: "user", content: "Hello!" }],
});
console.log(completion.choices[0].message.content);Model slug: anthropic/claude-3-haiku
Editor's verdict
Claude 3 Haiku is Anthropic's proprietary multimodal with a 200K-token context window.
On independent testing it scores 12.3 on the Artificial Analysis Intelligence Index.
At $1.25 per 1M output tokens, it is mid-priced for its class.
It is available through Anthropic's API and aggregators like OpenRouter.
Best suited to high speed and low latency and cost-effective for high-volume use.
Frequently asked questions
Claude 3 Haiku supports a context window of 200000 tokens.
User reviews
Real, verified reviews from the community shape this model's rating.
Loading reviews…
Other Claude models
Sibling versions in the Claude family from Anthropic.
Similar models
Other multimodal worth comparing.