Pharia-1-LLM-7B-control-aligned

Name: Pharia-1-LLM-7B-control-aligned
Author: Aleph Alpha

Catalogue only Open weights European provider

Sovereignty and hosting

Catalogue only

This model is catalogued for reference. It is not hosted on LLM EU yet, and no first-party endpoint serves it here. Requests to it go to the provider directly, under that provider's own terms.

Summary

Pharia-1-LLM-7B-control-aligned is Aleph Alpha’s 7B control model, published on Hugging Face under the Open Aleph License, which limits use to educational and research purposes. Aleph Alpha states the models were trained on curated data in compliance with EU and national copyright and data-protection rules and are optimised for German, French and Spanish. Commercial use runs through an on-premise agreement that includes the weights, not through the public download.

Specifications

Provider: Aleph Alpha (DE)
Context window: 8.2k
Parameters: 7.04B
Modalities: text
License: other · licence
Open weights: yes
Release date: 27 Aug 2024
Tasks: german, private, summarization, rag

Languages

Strong: de, fr, es

Adequate: en, it, pt, nl, pl

Languages

Strong

Adequate

Not verified: the provider has not documented this claim. — a language is listed as strong only where a source claims it; the card links the evidence where one exists.

When to use it

Strengths

German, French and Spanish optimisation with EU-focused data curation
control variant designed for length-controlled, non-refusing responses
weights downloadable for research and teaching
vendor offers commercial on-premise licensing of the same checkpoints

Limits

Open Aleph License allows non-commercial research and educational use only
8192-token context window
RLHF-style chat behaviour is limited compared with current instruct models

Pricing

This model is catalogued for reference. It is not hosted on LLM EU yet, and no first-party endpoint serves it here. Requests to it go to the provider directly, under that provider's own terms.

Verification status

Where a field has no source, the card shows it as unverified instead of filling it in. An empty field is information.

LLM EU Verified means documented, not legally certified. It is not an AI Act conformity assessment and LLM EU is not an EU body.

Versions and variants

catalogue · Catalogue only · weights

Example request

curl https://api.llmeu.com/v1/chat/completions \
  -H "Authorization: Bearer $LLMEU_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"pharia-1-llm-7b-control-aligned","messages":[{"role":"user","content":"Hello"}]}'

No playground is deployed for this model: no endpoint serves it, so a request would have nowhere to go.

Benchmarks

EU21 average (translated ARC, HellaSwag, TruthfulQA, MMLU): 0.417 (OpenGPT-X evaluation, listed as Pharia-1-LLM-7B-C-A) source