GPAI Ledger The public record of EU AI Act training-data summaries

GPAI LedgerOSAI index (open-source model releases) (European Open Source AI Index) › Capture 19 Aug 2026

OSAI index (open-source model releases) — capture 20260819T150753Z

ProviderEuropean Open Source AI Index
Targetwatched page — https://osai-index.eu/
Fetched (UTC)2026-08-19T15:07:52Z
Stored file7a8633bfc002ce522d41998a25cf4299ba0b4a448604c5a7569ebcdb659e4fb2.html.txt (866,239 bytes) (served with a .txt suffix so the captured page cannot run scripts on this site; bytes are identical — the SHA-256 verifies against this file)
SHA-2567a8633bfc002ce522d41998a25cf4299ba0b4a448604c5a7569ebcdb659e4fb2
OpenTimestamps proof7a8633bfc002ce522d41998a25cf4299ba0b4a448604c5a7569ebcdb659e4fb2.html.ots (calendar-attested; anchored in bitcoin over time)
WaybackWayback snapshot, 2026-08-19 15:07 UTC
Prior capture of this target— first capture of this target

Verify: sha256sum 7a8633bfc002ce522d41998a25cf4299ba0b4a448604c5a7569ebcdb659e4fb2.html.txt must equal the hash above (the filename IS the expected hash); ots verify 7a8633bfc002ce522d41998a25cf4299ba0b4a448604c5a7569ebcdb659e4fb2.html.ots -f 7a8633bfc002ce522d41998a25cf4299ba0b4a448604c5a7569ebcdb659e4fb2.html.txt (opentimestamps.org) proves the capture time (fresh proofs report 'pending' until bitcoin-anchored, typically within a day).

Extracted text

Machine-extracted text (layout may be lost; the authoritative content is the stored file above).

European Open Source AI Index
 European Open Source AI Index
 Menu
Database
News
Guides
About
Contribute

The index lists open generative AI models for
text
,
image
,
code
,
video
, &
audio
. Across the open-source AI landscape, we find large differences in actual openness.
 Explore all
0
 models ->
Last updated 03 July 2026
 Show filters
YuLan
 by Gaoling School of Artificial Intelligence
YuLan-Mini
 Compare
BLOOMZ
 by BigScience Workshop
BLOOM
 Compare
OLMo
 by Ai2
Olmo-3-1125-32B
 Compare
Apertus
 by Swiss AI Initiative
Apertus-70B-2509
 Compare
SmolLM
 by HuggingFace
SmolLM3-3B-Base
 Compare
mT0
 by BigScience Workshop
mT5-XXL
 Compare
Amber
 by LLM360
Amber
 Compare
Pythia
 by EleutherAI and Together Computer
Pythia-6.9B
 Compare
Open Assistant
 by Open Assistant
Pythia-12B
 Compare
Lucie
 by OpenLLM-France
Lucie-7B
 Compare
Instella
 by AMD
Instella-3B
 Compare
Arabic StableLM
 by StabilityAI
StableLM-2-1.6B
 Compare
K2
 by LLM360
K2-V2-Instruct
 Compare
EuroLLM
 by UTTER
EuroLLM-22B-2512
 Compare
Skywork-OR1
 by Skywork
DeepSeek-R1-Distill-Qwen-32B
 Compare
Omnilingual ASR
 by Meta
Omnilingual ASR
 Compare
MobileLLM
 by Meta
MobileLLM-R1-950M-base
 Compare
Fietje
 by Bram Vanroy
Phi-2
 Compare
Minerva
 by Sapienza Natural Language Processing Group
Minerva-7B-base-v1.0
 Compare
Whisper
 by OpenAI
Whisper-large-v3
 Compare
T5
 by Google AI
T5
 Compare
Eurus
 by OpenBMB
Mixtral-8x22B-v0.1
 Compare
Teuken
 by OpenGPT-X
Teuken-7B-base
 Compare
RedPajama
 by Together Computer
RedPajama-INCITE-7B-Base
 Compare
Poro
 by AMD Silo AI and TurkuNLP and High Performance Language Technologies (HPLT)
Llama-Poro-2-70B-Base
 Compare
OpenChat
 by OpenChat
Meta-Llama-3-8B
 Compare
Neo
 by Multimodal Art Projection
Neo-7B
 Compare
Guru
 by LLM360
Guru-32B
 Compare
CT-LLM
 by Multimodal Art Projection
CT-LLM-Base
 Compare
BTLM
 by Cerebras
BTLM-3B-8K-Base
 Compare
BERT
 by Google AI
BERT
 Compare
Vicuna
 by LMSYS
Vicuna-13B
 Compare
Tülu
 by Ai2
Llama-3.1-405B
 Compare
TildeOpen
 by Tilde.ai
TildeOpen-30b
 Compare
Salamandra
 by Barcelona Supercomputing Center
Salamandra-7B
 Compare
OpenMoE
 by Zheng Zian
OpenMoE-8B
 Compare
Occiglot
 by Occiglot
Occiglot-7B-EU5
 Compare
Llama Nemotron
 by NVIDIA
Llama-3.3-70B-Instruct
 Compare
GPT-SW3
 by AI Sweden
GPT-SW3-6.7B-V2
 Compare
GPT-NeoXT
 by Together Computer
GPT-NeoX-20B
 Compare
AquilaChat
 by Beijing Academy of Artificial Intelligence
Aquila2-70B-Expr
 Compare
Dolly
 by Databricks
Pythia-12B
 Compare
DeepHermes
 by Nous Research
Llama-3.1-70B
 Compare
Zephyr
 by HuggingFace
Mixtral-8x22B-v0.1
 Compare
WizardLM
 by Microsoft and Peking University
LLaMA-7B
 Compare
SynLogic
 by Minimax AI
SynLogic-32B
 Compare
Phi
 by Microsoft
Phi-4
 Compare
OpenELM
 by Apple
OpenELM-3B
 Compare
NeuralChat
 by Intel
Mistral-7B-v0.1
 Compare
Baguettotron
 by PleIAs
Baguettotron
 Compare
Pharia
 by Aleph Alpha Research
Pharia-1-LLM-7B
 Compare
minChatGPT
 by Ethan Yanjia Li
GPT2-Medium
 Compare
Yi
 by 01.AI
Yi-34B
 Compare
XBai-04
 by Yuan Shi Technology
Qwen3-32B
 Compare
StripedHyena
 by Together Computer
StripedHyena-Hessian-7B
 Compare
Saul
 by Equall
Mixtral-8x22B-v0.1
 Compare
Hunyuan
 by Tencent
Hunyuan-7B-Pretrain
 Compare
MiMo
 by Xiaomi
MiMo-V2-Flash-Base
 Compare
Xwin-LM
 by Xwin-LM
Llama-2-13B
 Compare
Geitje
 by Bram Vanroy
Mistral-7B-v0.1
 Compare
GPT OSS
 by OpenAI
unspecified
 Compare
Claire
 by OpenLLM-France
Falcon-7B
 Compare
BELLE
 by KE Technologies
Llama-2-13B
 Compare
Apriel
 by ServiceNow
Pixtral-12B
 Compare
UltraLM
 by OpenBMB
Llama-13B
 Compare
DeepSeek R1
 by DeepSeek
DeepSeek-V3-Base
 Compare
Solar
 by Upstage AI
unspecified
 Compare
Marco
 by Alibaba
Marco-LLM-GLO
 Compare
MPT
 by Databricks
MPT-30B
 Compare
Intern-S1
 by Shanghai AI Laboratory
Intern-S1-Pro
 Compare
Granite
 by IBM
Granite-4.0-H-Small-Base
 Compare
Viking
 by Silo AI and TurkuNLP and High Performance Language Technologies (HPLT)
unspecified
 Compare
Mistral NeMo
 by Mistral AI and NVIDIA
Mistral-NeMo-12B-Base
 Compare
DeepSeek V3.2
 by DeepSeek
DeepSeek-V3.1-Base
 Compare
Stanford Alpaca
 by Stanford University CRFM
Llama-7B
 Compare
Airoboros
 by Jon Durbin
Qwen1.5-110B
 Compare
Starling
 by NexusFlow
Llama-2-13B
 Compare
Seed-OSS
 by ByteDance
Seed-OSS-36B-Base
 Compare
Nanbeige
 by Nanbeige LLM lab
Nanbeige4-3B-Base
 Compare
LongAlign
 by Zhipu AI
Llama-2-13B
 Compare
Ling
 by Inclusion AI
Ling-2.5-1T
 Compare
Kimi K2.5
 by Moonshot AI
unspecified
 Compare
Gemma
 by Google AI
Gemma-3-27B-PT
 Compare
Falcon
 by Technology Innovation Institute
Falcon-H1-7B-Base
 Compare
Bielik
 by SpeakLeash AI
Bielik-11B-v3-Base-20250730
 Compare
Llama 4
 by Meta
Llama-4-Maverick-17B-128E
 Compare
Jamba
 by AI21
AI21-Jamba2-Mini
 Compare
GLM
 by Zhipu AI
GLM-5
 Compare
Llama 3.1
 by Meta
Llama-3.1-405B
 Compare
XVERSE
 by Shenzhen Yuanxiang Technology
XVERSE-MoE-A4.2B
 Compare
TeleChat
 by Tele-AI
unknown
 Compare
Snowflake Arctic
 by Snowflake
Snowflake-Arctic-Base
 Compare
RWKV
 by BlinkDL/RWKV
unspecified
 Compare
Persimmon
 by Adept AI Labs
Persimmon-8B-Base
 Compare
OPT
 by Meta
OPT-30B
 Compare
Mistral Large 3
 by Mistral AI
Mistral-Large-3-675B-Base-2512
 Compare
LFM2
 by Liquid AI
unknown
 Compare
Infinity-Instruct
 by Beijing Academy of Artificial Intelligence
Llama-3.1-70B
 Compare
H2O-Danube
 by H2O.ai
H2O-Danube3.1-4B-Base
 Compare
EXAONE
 by LG
unspecified
 Compare
BitNet
 by Microsoft
BitNet-b1.58-2B4T
 Compare
StableVicuna
 by CarperAI
LLaMA-13B
 Compare
Llama 2
 by Meta
Llama-2-70B
 Compare
LeoLM
 by LAION
Llama-2-70B
 Compare
Koala
 by BAIR
Llama-13B
 Compare
dots.llm1
 by RedNote
dots.llm1.base
 Compare
XGen
 by Salesforce
XGen-Small-9B-Base-R
 Compare
Reka Flash
 by Reka AI
unknown
 Compare
Jais
 by G42
unspecified
 Compare
FastChat-T5
 by LMSYS
Flan-T5-XL
 Compare
Crystal
 by LLM360
Crystal
 Compare
Llama 3.3
 by Meta
Llama-3.3-70B
 Compare
Stable Beluga
 by StabilityAI
Llama-2-70B
 Compare
Minimax-M2.5
 by Minimax AI
MiniMax-Text-01?
 Compare
Llama-Sherkala
 by G42
Llama-3.1-8B
 Compare
Cogito
 by DeepCogito
DeepSeek-V3-Base
 Compare
Llama 3
 by Meta
Meta-Llama-3-70B
 Compare
Qwen
 by Alibaba
unspecified
 Compare
Gemma Japanese
 by Google AI
Gemma-2-2B
 Compare
Baichuan
 by Baichuan Intelligent Technology
Baichuan2-13B-Base
 Compare
Command A
 by Cohere AI
Command A?
 Compare
Grok 2
 by xAI
unknown
 Compare
Lumo AI
 by Proton
Undisclosed
 Compare
 Explore all
0
 models ->
Latest news
02 December 2025
Mistral 3: A Heavyweight Model, Lightweight on Openness
The Mistral 3 family, including the massive 675B Mistral Large 3, fills a gap for high-performance European models and aids local data governance, it fails to set a new standard for openness.
29 August 2025
Lumo: the least open 'open' AI assistant
Proton sets a new record in open-washing
01 August 2025
BERT: The original sin of open-washing large language models
Already in 2018 Google's pioneering LLM 'BERT' was marketed as open source, leading an entire field down a dubious path.
18 July 2025
MusGO: a new openness index for the music domain
A new resource building on our openness framework
28 May 2025
YuLan-Mini: A new highly open model
Gaoling School of Artificial Intelligence releases a new highly open AI model
27 May 2025
OpenGPT-X: Open Source AI 'Made in Germany' falls short of own claims
In this deep dive we evaluate OpenGPT-X's flagship Teuken models through the lens of openness
03 April 2025
Introducing Performance Classes in the OS AI Index
How we classify models under 3 performance classes: Limited, Full and Latest
24 February 2025
Introducing the European Open Source AI Index
In which we introduce the European Open Source AI Index and explain how to follow a moving target
17 February 2025
Luminous: the language model that wasn't
In which we trace the silent demise of a hyped German large language model
25 January 2025
How Open is DeepSeek?
Investigating the open-source status of DeepSeek-R1
Read all news ->
As seen in
Partners
Supported by the Centre for Language Studies and the Dutch Research Council. Website design & development © 2024 by
BSTN
. This version of the index generated
03 July 2026
, website content last updated
10 July 2026
.