OpenAssistant Conversations (OASST1)
OpenAssistantCrowdsourced human dialogue corpus from the 2023 OpenAssistant project, holding about 161,000 messages across 35 languages contributed by the LAION community. It remains a staple human-data ingredient in open post-training mixes, from Guanaco through Tulu 3 to OLMo 3's Dolci.
Openness
5 high confidence- license
- Apache-2.0
- gated
- false
- dataset_card
- present
Apache-2.0, ungated, full dataset card.
- https://huggingface.co/api/datasets/OpenAssistant/oasst1 recorded 2026-09-29
Hub API record shows the dataset is public and ungated with tag "license:apache-2.0" and a description opening "OpenAssistant Conversations Dataset (OASST1)".
- https://huggingface.co/datasets/OpenAssistant/oasst1 recorded 2026-09-29
The Hugging Face dataset page renders the card for OpenAssistant/oasst1 with a public, ungated file listing; it carries "OpenAssistant Conversations Dataset (OASST1)" and the tag "license:apache-2.0".
- https://huggingface.co/datasets/OpenAssistant/oasst1/raw/main/README.md recorded 2026-09-29
Raw dataset card front matter declares "license: apache-2.0" and the summary describes "a human-generated, human-annotated assistant-style conversation corpus consisting of 161,443 messages in 35 different languages".
Adoption
3 high confidenceAdoption is measured by Hugging Face downloads of the dataset over the trailing month, the only usage channel on record for OASST1.
- https://huggingface.co/api/datasets/OpenAssistant/oasst1 recorded 2026-08-13
20,967 downloads in the trailing 30 days for OpenAssistant/oasst1
Capability
3 high confidenceOASST1 still appears in later post-training mixtures such as Guanaco, Tulu 3 and OLMo 3's Dolci, but only as a small fraction of a larger blend, with no benchmark isolating its own contribution. The QLoRA paper fine-tuned Guanaco 65B on a 9,209-example variant of OASST1 and called the result the best-performing open-source chatbot model of its time.
- https://arxiv.org/html/2305.14314 recorded 2026-09-29
QLoRA paper full text says "Guanaco 65B, which we finetune on a variant of OASST1, is the best-performing open-source chatbot model and offers performance competitive to ChatGPT" and, on the dataset, "we only use the top reply at each level in the conversation tree. This limits the dataset to 9,209 examples."
- https://huggingface.co/datasets/allenai/Dolci-Instruct-SFT/raw/main/README.md recorded 2026-09-29
Dolci Instruct SFT card lists "[OpenAssistant Guanaco](https://huggingface.co/datasets/OpenAssistant/oasst1) (Apache 2.0), 7,132 prompts" among its components.
- https://huggingface.co/datasets/allenai/tulu-3-sft-mixture/raw/main/README.md recorded 2026-09-29
Tulu 3 SFT mixture card lists "[OpenAssistant Guanaco](https://huggingface.co/datasets/OpenAssistant/oasst1) (Apache 2.0), 7,132 prompts (Kopf et al., 2024)" among its components.
Verified 2026-09-27