Project AURIONAI Use-Cases Research & Innovation
Google
LLMfamily

Flan-T5

T5 checkpoints instruction-tuned on a large task collection (1.8K tasks in the paper), giving better zero-shot and few-shot behaviour for question answering and summarisation.

GoogleGoogleUSA
Diagram of Flan instruction finetuning across many tasks, with example prompts and answers, from the Flan-T5 model card.
Levelfamily
ClassLLM
Parametersnot in the notes
Contextnot in the notes
Released20 Oct 2022
Opennessopen weights permissive
Runs onnot in the notes
Statusactive

Lineage

20222023Flan-T5 baseFlan-T5 base, Oct 2022Flan-T5 large, Oct 2022Flan-T5 XXL, Oct 2022
GoogleFlan-T5 base belongs to this familyLLM21 Oct 2022247.58M
GoogleFlan-T5 large belongs to this familyLLM21 Oct 2022783.15M
GoogleFlan-T5 XXL belongs to this familyLLM21 Oct 202211.27B
Where this comes from
Parent and child relations from model cards, Hugging Face base_model fields and the genealogy tables of the research notes. Confidence and evidence are the notes' own.

R&D Topics

Open post-training recipesexemplarText-to-text pretraining and summarisationexemplar

Identity

Developer
Google
Family
not in the notes
Released
20 Oct 2022 sourcesannouncement page
Weights published
–
Record checked
12 Oct 2026
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.

Architecture and Size

Architecture
dense transformer
Total parameters
– sources
Active parameters
–
Checkpoint parameters
–
Context length
not in the notes
Input
text
Output
text
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).

Openness and Access

Openness
open weights permissive
Licence
Apache License 2.0 sources
Access routes
open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.

Deployment

Runs on
not in the notes
Footprint
not in the notes
Runtimes
not in the notes
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.

Indonesia

Indonesian support
not stated
Regional languages
not in the notes
Jakarta hosting
unknown
Languages
Model card languages: en, fr, ro, de, multilingual
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.

Benchmark Results

The notes record no benchmark result for this model.

Industry Use Cases

No industry use case links this model.

Sources

  1. arxiv.org/abs/2210.11416 https://arxiv.org/abs/2210.11416
  2. Model Garden on Gemini Enterprise Agent Platform cloud.google.com https://cloud.google.com/model-garden
  3. huggingface.co/api/models?pipeline_tag=question-answering&so... https://huggingface.co/api/models?pipeline_tag=question-answering&sort=downloads&limit=50
  4. huggingface.co/google/flan-t5-base https://huggingface.co/google/flan-t5-base
  5. huggingface.co/google/flan-t5-large https://huggingface.co/google/flan-t5-large

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Flan-T5", https://aurion.semai.id/models/google-flan-t5/, records of October 2026.