MM1-30B
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
- Apple
- Organisation type
- Industry
- Country
- United States of America
- Published
- 14 March 2024
- Authors
- Brandon McKinzie, Zhe Gan, Jean-Philippe Fauconnier, Sam Dodge, Bowen Zhang, Philipp Dufter, Dhruti Shah, Xianzhi Du, Futang Peng, Floris Weers, Anton Belyi, Haotian Zhang, Karanjeet Singh, Doug Kang, Ankur Jain, Hongyu Hè, Max Schwarzer, Tom Gunter, Xiang Kong, Aonan Zhang, Jianyu Wang, Chong Wang, Nan Du, Tao Lei, Sam Wiseman, Guoli Yin, Mark Lee, Zirui Wang, Ruoming Pang, Peter Grasch, Alexande…
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Multimodal, Language, Vision
- Task
- Chat, Image captioning, Visual question answering
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
- 30B
- Training data
- tokens
30B
at least 2T tokens
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
- 4.9 × 10²³ FLOP
- How it was established
- Operation counting
Pre-trained on ~2B image-text pairs and 2T tokens (Table 2). Each image is 144 tokens, so the images are ~300B tokens. Then additional multimodal training for 400B tokens, for a total of ~2.7T tokens. This is the final training recipe: "We initialize both the image encoder and the underlying LLM decoder weights for MM1 from in-house pre-trained models2. We then perform multimodal pre-training on the above data mix for 200k steps (approx. 400B tokens)." Compute = 6ND = 6 * 2.7 trillion * 30 bi…
Availability
Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.
- Weights
- Closed — provider access only
- Model access
- Unreleased
- Training code
- Unreleased
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Likely above 10²³ FLOP
- Yes
- Why it is tracked
- SOTA improvement
- Record confidence
- Likely
- Citations
- 266
"In particular, the pretrained model MM1 is SOTA, performing better than Emu2 [105], Flamingo [3], and IDEFICS [47] on captioning and visual question answering (VQA) tasks in few-shot settings, both in small and large size regimes" Table 4: outperforms Gemini and GPT-4V on VQA
Sources
Where this record came from and when it was last checked.
- Reference
- MM1: Methods, Analysis & Insights from Multimodal LLM Pre-training
- Last updated
- 25 May 2026
What the numbers mean
Background
MM1-30B was published by Apple, in the country recorded as United States of America, during March 2024. The publishing organisation is categorised as industry.
It works in the domain of Multimodal, Language, Vision, and is recorded as performing the task of chat, Image captioning, Visual question answering.
Because the weights are not available, none of the hardware figures elsewhere on this site apply to it.
What went into building it
Producing it required arithmetic totalling around 4.9 × 10²³ FLOP. That figure measures what producing the model cost, and has no bearing on how fast it answers.
The reason it appears in this catalogue at all: sOTA improvement.
Answers
MM1-30B — common questions
MM1-30B— how much compute was used to train it?
Training consumed around 4.9 × 10²³ 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.
MM1-30B— what GPU do I need to run it?
None. This 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.
MM1-30B— is it open source?
No. Its weights have not been published, so it exists only as a service controlled by its owner.
MM1-30B— how many parameters does it have?
It has a parameter count of 30B. 30B. 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.
MM1-30B— who created it?
It was published by Apple, based in United States of America, an organisation categorised as industry.
MM1-30B— when was it released?
It was published in March 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.
MM1-30B— what is it used for?
It works in the domain of Multimodal, Language, Vision, and is recorded as handling the task of chat, Image captioning, Visual question answering. These are the areas it was designed around; they describe intent rather than a hard boundary.
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.