Anthropic · Efficient
Claude Haiku Latest
Compatibility alias that routes to the newest version of Claude Haiku. Currently routes to Claude Haiku 5.5 with adaptive thinking.
- Reasoning
- Tools
- Vision
- Structured output
- PDF input
- Input
- Text, Images
- Output
- Text
- Reasoning effort
- low · medium · high · xhigh · max
- Discovery tier
- Efficient
Pricing
USD per 1M tokens- Input
- $0.10
- Output
- $0.50
- Cache read
- $0.01
- Cache write
- $0.13
- $0.20 for a 1-hour cache
Reference prices as published by the model provider. API example
Specifications
- Context window
- 1M
- Max output
- 128K
- Average output (7 days)
- 832 tokens
- Longer than 66% of 465 models
Side by side
Compare models
Put Claude Haiku Latest next to up to three other models. The comparison stays in the link, so you can share it.
Public evaluations
Benchmarks
Independent evaluation results, with how each score ranks among the models in the Noumen catalog.
Intelligence Index
43.4
Tested at max reasoning effort
Better than 88% of 326 scored models
Agentic work
AutomationBench-AA
35.4%Workflow automation, penalized for guardrail violations
Better than 50% of 173 scored models
AutomationBench-AA tasks completed
13.9%Workflows finished end to end without a guardrail violation
Better than 57% of 173 scored models
Harvey LAB-AAv1.1
0.6%Legal tasks passed in full, with no material hallucination
Better than 62% of 29 scored models
AA-Briefcasev1.1
1,575 EloAgentic knowledge work, rated head to head
Better than 90% of 166 scored models
GDPval-AAv2.1
1,619 EloEconomically valuable occupational tasks, rated head to head
Better than 88% of 187 scored models
Document reasoning
GDP.pdf
20.8%Professional reasoning over PDF documents, all-pass rate
Better than 77% of 171 scored models
AA-LCRv1.1
82.7%Reasoning across long documents
Better than 88% of 309 scored models
MLCR-AA
35.0%Long-context reasoning over medical records
Better than 96% of 121 scored models
Reasoning and knowledge
Humanity’s Last Exam
44.4%Expert-level questions across academic fields
Better than 86% of 323 scored models
Coding
Terminal-Benchv4.0
32.8%Practical software engineering in a terminal
Better than 78% of 171 scored models
SciCode
55.0%Scientific computing in Python
Better than 74% of 174 scored models
Legacy evaluations
AA-LCR (unversioned)
82.7%Long-context reasoning, earlier grading
Better than 88% of 309 scored models
Scores from Artificial Analysis, updated Oct 26, 2026. Claude Haiku Latest was tested at max reasoning effort; compare scores measured at the same effort.Artificial Analysis
Live routing data
Performance and reliability
Recent averages across requests routed for this model. Your own speed depends on prompt length, reasoning effort and load.
- Throughput
- 193.9 tok/s
- Faster than 85% of 250 models
- Time to first token
- 4.9 s
- Quicker than 34% of 250 models
- Prompt cache hits
- 0.0%
- About $0.10 per 1M input tokens with caching
- Request success (24h)
- 99.9%
- Reliable
Cache hits are the share of input tokens recently served from a prompt cache across all requests. They are not a forecast of your own hit rate.
- Healthy (98% or more)
- Degraded (90–98%)
- Down (under 90%)
- No requests
Developers
Use it through the API
Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.
curl https://www.noumen.xyz/api/v1/chat/completions \
-H "Authorization: Bearer $NOUMEN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "anthropic/claude-haiku-latest",
"reasoning_effort": "medium",
"messages": [{ "role": "user", "content": "Explain this stack trace." }]
}'import os
from openai import OpenAI
client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
model="anthropic/claude-haiku-latest",
messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)Anthropic
Related models
Other Anthropic models, starting with the same family.