Their EU contact
If something goes wrong inside the EU, this is who to contact.
Anthropic Ireland, Limited
6th Floor South Bank House, Barrow Street, Dublin, Leinster, IE
Back to Directory
TypeAI Model
Made byAnthropic
FromUS
Last checkedMay 5, 2026Added May 5, 2026
Claude 3 Haiku
by Anthropic
AI ModelNot fully checked yet
Claude 3 Haiku is the fastest and most cost-efficient model in the Claude 3 family, released by Anthropic in March 2024. It is optimized for low-latency, high-volume tasks like real-time chat, lightweight agentic workflows, and simple reasoning. Despite its smaller size, it delivers surprisingly strong performance for its class. Claude 3 Haiku was trained on the same proprietary mix of data as Claude 3 Opus and Sonnet (public internet data up to August 2023, private datasets, and internally generated data). It underwent RLHF and Constitutional AI fine-tuning to align with helpfulness, honesty, and harmlessness. The model is multimodal (text and vision) and multilingual. Safety evaluations show strong alignment, with low rates of harmful outputs and high resistance to jailbreaks. It was deployed under AI Safety Level 2 (ASL-2) protections.
- 01
Browse
Built by Anthropic in US. Your data passes through them.
- 02
Track
We watch for changes to their terms so you don't have to.
- 03
Stay in control
1 thing to think about before letting it touch your work
How open is this maker?
Other companies
20/20
Following rules
10/20
Documents
10/25
How it works
15/25
Care & caution
5/10
60%
Pretty open openness
02
Openness Assessment
18%
Closed
Availability
0/5
Documentation
1/6
Access Methods
2/3
03Who built it
Who's behind this AI
The people and company behind this AI
Legal name
Anthropic PBC
Also goes by: Anthropic
Website
anthropic.comEmail
press@anthropic.comOffice address
548 Market Street, San Fransisco, CA, 94105, US
Where they're based
Their headquarters
Where the company is officially based.
Headquarters
04Legal & paperwork
What the maker has told EU regulators
What this maker has officially told EU regulators about how their AI works.
Click to see compliance details
These four sections are required by the EU AI Act. Auditors and compliance teams use them — feel free to skim.
How well we know thisUnverified
Stamps & certificates
Stamps and certificates that say the AI passed EU checks.
- PendingEU safety check (CE marking)
- How to use it (official guide)
Where in Europe you can get it
Which EU countries this AI is sold in.
What they've told the public
What the maker has said publicly about how the AI works.
05What to watch for
Things to think about before using it
Potential Risks
1 considerations identified
Review recommended before use
1High
06Other companies involved
Other companies that may touch your data
Other companies this maker works with that may end up handling your data.
07The fine print
Their legal & privacy documents
The legal documents — terms, privacy, security — that spell out the rules.
Terms & legal
Privacy & security
08EU-based
Made in Europe
Discover EU-based alternatives for this AI application.
01
AI Model02AI Model03AI Model04AI Model
Browse all EU alternativesMinistral 3 14B Instruct 2512· Mistral AI
Mistral Medium· Mistral AI
Magistral Medium 1.0· Mistral AI
Mistral Large 2· Mistral AI