Claude Haiku 3.5 - In-Depth Overview

Anthropic · Claude 3.5

Current

Latest in family

Model ID: claude-3-5-haiku-20241022

Get all the details on Claude Haiku 3.5, an AI model from Anthropic. This page covers its token limits, pricing structure, key capabilities such as prompt_caching, streaming_messages, batch_processing, available API code samples, and performance strengths.

Key Metrics

Input Limit

200K tokens

Output Limit

8.2K tokens

Input Cost

$0.80/1M

Output Cost

$4.00/1M

Sample API Code

from anthropic import Anthropic
client = Anthropic()
message = client.messages.create(
    model="claude-3-5-haiku-20241022",
    max_tokens=1024,
    messages=[
        {"role": "user", "content": "Hello, Claude!"}
    ]
)
print(message.content)

Required Libraries

anthropic

@anthropic-ai/sdk

Benchmarks

Benchmark	Score	Source	Notes
lmarena text	1234	OpenLLM Leaderboard	Rank 77
lmarena webdev	1133	OpenLLM Leaderboard	Rank 13
lmarena vision	1000	OpenLLM Leaderboard	Rank 50
livebench global average	44.98	LiveBench	-
livebench reasoning average	26.19	LiveBench	-
livebench coding average	53.17	LiveBench	-
livebench mathematics average	34.84	LiveBench	-
livebench data analysis average	54.12	LiveBench	-
livebench language average	39.71	LiveBench	-
livebench if average	61.88	LiveBench	-
gpqa	41.6%	Vellum	-
swe bench	40.6%	Vellum	-
math 500	69.4%	Vellum	-
bfcl	54.31%	Vellum	-
alder polyglot	28%	Vellum	-