xGen-MM (BLIP-3)
No estimate
No hardware requirements for this model
The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.
On record
Full specification
Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.
Origin
Who built this model, where, and when it was published.
- Organisation
- Salesforce Research,University of Washington
- Organisation type
- Industry,Academia
- Country
- United States of America
- Published
- 16 August 2024
- Authors
- Le Xue, Manli Shu, Anas Awadalla, Jun Wang, An Yan, Senthil Purushwalkam, Honglu Zhou, Viraj Prabhu, Yutong Dai, Michael S Ryoo, Shrikant Kendre, Jieyu Zhang, Can Qin, Shu Zhang, Chia-Chih Chen, Ning Yu, Juntao Tan, Tulika Manoj Awalgaonkar, Shelby Heinecke, Huan Wang, Yejin Choi, Ludwig Schmidt, Zeyuan Chen, Silvio Savarese, Juan Carlos Niebles, Caiming Xiong, Ran Xu
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Multimodal, Vision, Language
- Task
- Image captioning, Character recognition (OCR), Visual question answering, Chat
- Base model
- phi-3-mini 3.8B
Size
How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.
- Parameters
- 4B
- Training data
- tokens
Table 1: shows the model is called xGen-MM-base (4B)
"About 100B multimodal tokens" for pretraining SFT phase: "Ultimately, we collect a mixture of 1 million publically available instruction-tuning samples, on which we fine-tune our model for one epoch." DPO phase: "We thus generate 62.6k preference examples" LoRA phase Safety Fine-tuning phase: "2k examples of unsafe images and instructions...5k additional examples from the instruction fine-tuning dataset" Total is about 100 billion.
Training compute
The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.
- Training compute
- 2.4 × 10²¹ FLOP
Assuming 1 epoch and dense architecture (phi-3 mini is a dense model), compute = 6 * 4B * 100B = 2.4e+21 FLOP.
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Record confidence
- Confident
- Citations
- 139
Sources
Where this record came from and when it was last checked.
- Reference
- xGen-MM (BLIP-3): A Family of Open Large Multimodal Models
- Last updated
- 11 February 2026
What the numbers mean
Background
xGen-MM (BLIP-3) was published by Salesforce Research,University of Washington, in United States of America, in August 2024. It comes out of industry,Academia.
It works in Multimodal, Vision, Language, and is recorded as doing image captioning, Character recognition (OCR), Visual question answering, Chat.
Its starting point was phi-3-mini 3.8B — most models at this scale are adapted from an existing base rather than built from nothing.
This is a closed model: the trained values stayed with whoever produced them, and there is no local version to run.
How it was trained
Training it took roughly 2.4 × 10²¹ FLOP of computation — a measure of what producing the model cost, not of how fast it answers.
Answers
xGen-MM (BLIP-3) — common questions
Is xGen-MM (BLIP-3) open source?
The licensing for xGen-MM (BLIP-3) was never recorded in our source data. We treat unstated licensing as closed, because an unrecorded licence is not one to rely on.
How many parameters does xGen-MM (BLIP-3) have?
xGen-MM (BLIP-3) has 4B parameters. Table 1: shows the model is called xGen-MM-base (4B). That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.
Who created xGen-MM (BLIP-3)?
xGen-MM (BLIP-3) was published by Salesforce Research,University of Washington, based in United States of America, categorised as industry,Academia.
When was xGen-MM (BLIP-3) released?
xGen-MM (BLIP-3) was published in August 2024. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.
What is xGen-MM (BLIP-3) used for?
xGen-MM (BLIP-3) works in Multimodal, Vision, Language, and is recorded as handling image captioning, Character recognition (OCR), Visual question answering, Chat. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.
How much compute was used to train xGen-MM (BLIP-3)?
Around 2.4 × 10²¹ FLOP. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.
What GPU do I need to run xGen-MM (BLIP-3)?
None. xGen-MM (BLIP-3) is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.
The other direction
Looking at it from the other side?
This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.