Florence-2-base-ft How much memory it needs, what its licence allows, and whether it fits an office box.
Short answer. Florence-2-base-ft needs about 1.6 GB of memory at 4-bit, 1.7 GB at 8-bit and 2.0 GB at full 16-bit precision, including runtime overhead. It has about 232 million parameters and a 1,024-token context window. Licence in short: commercial use allowed. By memory alone, on ANVE’s launch configurations: ANVE One fits at BF16, ANVE Office fits at BF16, ANVE Office Pro (GPU) fits at BF16. These are estimates from its published size, not benchmarks.
01
What it is
Florence-2-base-ft is a document-reading model that turns scanned pages and images into text, published by Microsoft on Hugging Face on 15 June 2024. It has about 232 million parameters and a configured context window of 1,024 tokens.
Its card lists the licence as MIT. In short, the licence allows commercial use, modification and redistribution as long as the copyright notice is kept.
Hugging Face counted about 191,298 downloads of it in the 30 days before we checked (7 Oct 2026), and 148 people have liked it there.
02
How much memory it needs
The arithmetic: parameters × bits per weight ÷ 8 gives the weights, plus about 1.5 GB for the runtime. These are planning estimates, not measurements.
Inputs are short, so working memory is small next to the weights.
Precision
Weights
Total, estimated
BF16 (full precision)
0.46 GB
2.0 GB
8-bit (Q8_0)
0.25 GB
1.7 GB
4-bit (Q4_K)
0.13 GB
1.6 GB
03
Architecture at a glance
Setting
Value in config.json
Architecture class
Florence2ForConditionalGeneration
Model type
florence2
Layers
6
Vocabulary
51,289
Configured context
1,024
04
Will it fit on ANVE?
These verdicts are arithmetic, not tests, against the specification at launch: ANVE One can give a model about three quarters of its 64 GB of unified memory, while ANVE Office and ANVE Office Pro hold it on a 24 GB graphics card and keep the rest of the stack in 64 GB or 128 GB of system memory. ANVE’s tested default is Qwen 3.8 27B; this model is not yet validated on ANVE.
Configuration (at launch)
Memory for the model
Verdict
ANVE One
about 48 GB of 64 GB unified memory
fits at BF16
ANVE Office
24 GB on the RTX PRO 4000 Blackwell SFF Edition
fits at BF16
ANVE Office Pro (GPU)
24 GB on the RTX PRO 4000 Blackwell
fits at BF16
05
What an office would use it for
For an office, this is how scanned invoices, letters and forms become searchable text. Check its output on your own documents, especially tables, stamps and handwriting.
06
Similar models to compare
granite-docling-258M (0.3B parameters, about 1.6 GB at 4-bit before context; commercial use allowed)
Nemotron-Parse-2.0 (0.9B parameters, about 2.0 GB at 4-bit before context; custom licence: check before commercial use)
Nemotron-Parse-v1.1 (1.0B parameters, about 2.0 GB at 4-bit before context; custom licence: check before commercial use)
07
Questions people ask
How much memory does Florence-2-base-ft need?
About 1.6 GB at 4-bit, 1.7 GB at 8-bit and 2.0 GB at full 16-bit precision, including about 1.5 GB of runtime overhead. These are estimates from its parameter count.
Can Florence-2-base-ft run on a 24 GB graphics card?
Yes, at BF16 by this estimate, with room left for context.
Is Florence-2-base-ft free for commercial use?
Its card lists MIT, which allows commercial use, modification and redistribution as long as the copyright notice is kept. Read the licence itself before relying on this summary; it is not legal advice.
What is Florence-2-base-ft’s context window?
Its configuration sets 1,024 tokens. Longer contexts need more memory for the KV cache, so the practical limit on a given machine can be lower.
08
What is not real yet
No sealed ANVE unit exists; every image of the appliance on this site is a labelled render. The software runs today on a development machine. No customers yet, no certifications, no delivery date, and the final price is not announced.
Reservations are ₹999 in India or US$12 elsewhere, refundable in full until you confirm an order in writing (terms). Offices that cannot use cloud AI can apply to be one of ten design partners.
