Ovis-Image Technical Report Article Swipe

PDF

YOU? · · 2025 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2511.22982

We introduce $\textbf{Ovis-Image}$, a 7B text-to-image model specifically optimized for high-quality text rendering, designed to operate efficiently under stringent computational constraints. Built upon our previous Ovis-U1 framework, Ovis-Image integrates a diffusion-based visual decoder with the stronger Ovis 2.5 multimodal backbone, leveraging a text-centric training pipeline that combines large-scale pre-training with carefully tailored post-training refinements. Despite its compact architecture, Ovis-Image achieves text rendering performance on par with significantly larger open models such as Qwen-Image and approaches closed-source systems like Seedream and GPT4o. Crucially, the model remains deployable on a single high-end GPU with moderate memory, narrowing the gap between frontier-level text rendering and practical deployment. Our results indicate that combining a strong multimodal backbone with a carefully designed, text-focused training recipe is sufficient to achieve reliable bilingual text rendering without resorting to oversized or proprietary models.

Related Topics

Minority Report (Film)

World Happiness Report

Drudge Report

Weather Report

The Report (2019 Film)

Technical Ecstasy

The Majority Report With Sam Seder

Technical University Of Munich

U.S. News & World Report Best Colleges Ranking

Xbox 360 Technical Problems

Technical (Vehicle)

Playstation 3 Technical Specifications

Concepts

No concepts available.

Metadata

Type: preprint
Landing Page: http://arxiv.org/abs/2511.22982
PDF: https://arxiv.org/pdf/2511.22982
OA Status: green
OpenAlex ID: https://openalex.org/W4416951193

All OpenAlex metadata

Raw OpenAlex JSON

OpenAlex ID: https://openalex.org/W4416951193

Canonical identifier for this work in OpenAlex
DOI: https://doi.org/10.48550/arxiv.2511.22982

Digital Object Identifier
Title: Ovis-Image Technical Report

Work title
Type: preprint

OpenAlex work type
Publication year: 2025

Year of publication
Publication date: 2025-11-28

Full publication date if available
Authors: Liangfu Cao, M.W. Fu, Lan Li

List of authors in order
Landing page: https://arxiv.org/abs/2511.22982

Publisher landing page
PDF URL: https://arxiv.org/pdf/2511.22982

Direct link to full text PDF
Open access: Yes

Whether a free full text is available
OA status: green

Open access status per OpenAlex
OA URL: https://arxiv.org/pdf/2511.22982

Direct OA link when available
Cited by: 0

Total citation count in OpenAlex

Full payload

id	https://openalex.org/W4416951193
doi	https://doi.org/10.48550/arxiv.2511.22982
ids.doi	https://doi.org/10.48550/arxiv.2511.22982
ids.openalex	https://openalex.org/W4416951193
fwci
type	preprint
title	Ovis-Image Technical Report
biblio.issue
biblio.volume
biblio.last_page
biblio.first_page
is_xpac	False
apc_list
apc_paid
language
locations[0].id	pmh:oai:arXiv.org:2511.22982
locations[0].is_oa	True
locations[0].source.id	https://openalex.org/S4306400194
locations[0].source.issn
locations[0].source.type	repository
locations[0].source.is_oa	True
locations[0].source.issn_l
locations[0].source.is_core	False
locations[0].source.is_in_doaj	False
locations[0].source.display_name	arXiv (Cornell University)
locations[0].source.host_organization	https://openalex.org/I205783295
locations[0].source.host_organization_name	Cornell University
locations[0].source.host_organization_lineage	https://openalex.org/I205783295
locations[0].license
locations[0].pdf_url	https://arxiv.org/pdf/2511.22982
locations[0].version	submittedVersion
locations[0].raw_type	text
locations[0].license_id
locations[0].is_accepted	False
locations[0].is_published	False
locations[0].raw_source_name
locations[0].landing_page_url	http://arxiv.org/abs/2511.22982
locations[1].id	doi:10.48550/arxiv.2511.22982
locations[1].is_oa	True
locations[1].source.id	https://openalex.org/S4306400194
locations[1].source.issn
locations[1].source.type	repository
locations[1].source.is_oa	True
locations[1].source.issn_l
locations[1].source.is_core	False
locations[1].source.is_in_doaj	False
locations[1].source.display_name	arXiv (Cornell University)
locations[1].source.host_organization	https://openalex.org/I205783295
locations[1].source.host_organization_name	Cornell University
locations[1].source.host_organization_lineage	https://openalex.org/I205783295
locations[1].license
locations[1].pdf_url
locations[1].version
locations[1].raw_type	article
locations[1].license_id
locations[1].is_accepted	False
locations[1].is_published
locations[1].raw_source_name
locations[1].landing_page_url	https://doi.org/10.48550/arxiv.2511.22982
indexed_in	arxiv, datacite
authorships[0].author.id	https://openalex.org/A5080221188
authorships[0].author.orcid
authorships[0].author.display_name	Liangfu Cao
authorships[0].author_position	middle
authorships[0].raw_author_name	Cao, Liangfu
authorships[0].is_corresponding	False
authorships[1].author.id	https://openalex.org/A5072631669
authorships[1].author.orcid	https://orcid.org/0000-0002-0409-9701
authorships[1].author.display_name	M.W. Fu
authorships[1].author_position	middle
authorships[1].raw_author_name	Fu, Minghao
authorships[1].is_corresponding	False
authorships[2].author.id	https://openalex.org/A5021251089
authorships[2].author.orcid	https://orcid.org/0000-0001-5698-1045
authorships[2].author.display_name	Lan Li
authorships[2].author_position	middle
authorships[2].raw_author_name	Li, Lan
authorships[2].is_corresponding	False
has_content.pdf	False
has_content.grobid_xml	False
is_paratext	False
open_access.is_oa	True
open_access.oa_url	https://arxiv.org/pdf/2511.22982
open_access.oa_status	green
open_access.any_repository_has_fulltext	False
created_date	2025-12-03T00:00:00
display_name	Ovis-Image Technical Report
has_fulltext	False
is_retracted	False
updated_date	2025-12-03T14:19:12.882513
primary_topic
cited_by_count	0
locations_count	2
best_oa_location.id	pmh:oai:arXiv.org:2511.22982
best_oa_location.is_oa	True
best_oa_location.source.id	https://openalex.org/S4306400194
best_oa_location.source.issn
best_oa_location.source.type	repository
best_oa_location.source.is_oa	True
best_oa_location.source.issn_l
best_oa_location.source.is_core	False
best_oa_location.source.is_in_doaj	False
best_oa_location.source.display_name	arXiv (Cornell University)
best_oa_location.source.host_organization	https://openalex.org/I205783295
best_oa_location.source.host_organization_name	Cornell University
best_oa_location.source.host_organization_lineage	https://openalex.org/I205783295
best_oa_location.license
best_oa_location.pdf_url	https://arxiv.org/pdf/2511.22982
best_oa_location.version	submittedVersion
best_oa_location.raw_type	text
best_oa_location.license_id
best_oa_location.is_accepted	False
best_oa_location.is_published	False
best_oa_location.raw_source_name
best_oa_location.landing_page_url	http://arxiv.org/abs/2511.22982
primary_location.id	pmh:oai:arXiv.org:2511.22982
primary_location.is_oa	True
primary_location.source.id	https://openalex.org/S4306400194
primary_location.source.issn
primary_location.source.type	repository
primary_location.source.is_oa	True
primary_location.source.issn_l
primary_location.source.is_core	False
primary_location.source.is_in_doaj	False
primary_location.source.display_name	arXiv (Cornell University)
primary_location.source.host_organization	https://openalex.org/I205783295
primary_location.source.host_organization_name	Cornell University
primary_location.source.host_organization_lineage	https://openalex.org/I205783295
primary_location.license
primary_location.pdf_url	https://arxiv.org/pdf/2511.22982
primary_location.version	submittedVersion
primary_location.raw_type	text
primary_location.license_id
primary_location.is_accepted	False
primary_location.is_published	False
primary_location.raw_source_name
primary_location.landing_page_url	http://arxiv.org/abs/2511.22982
publication_date	2025-11-28
publication_year	2025
referenced_works_count	0
abstract_inverted_index.a	3, 29, 41, 87, 109, 114
abstract_inverted_index.7B	4
abstract_inverted_index.We	0
abstract_inverted_index.as	71
abstract_inverted_index.is	120
abstract_inverted_index.on	63, 86
abstract_inverted_index.or	132
abstract_inverted_index.to	14, 122, 130
abstract_inverted_index.2.5	37
abstract_inverted_index.GPU	90
abstract_inverted_index.Our	104
abstract_inverted_index.and	73, 79, 101
abstract_inverted_index.for	9
abstract_inverted_index.gap	96
abstract_inverted_index.its	55
abstract_inverted_index.our	23
abstract_inverted_index.par	64
abstract_inverted_index.the	34, 82, 95
abstract_inverted_index.Ovis	36
abstract_inverted_index.like	77
abstract_inverted_index.open	68
abstract_inverted_index.such	70
abstract_inverted_index.text	11, 60, 99, 126
abstract_inverted_index.that	45, 107
abstract_inverted_index.upon	22
abstract_inverted_index.with	33, 49, 65, 91, 113
abstract_inverted_index.Built	21
abstract_inverted_index.model	6, 83
abstract_inverted_index.under	17
abstract_inverted_index.GPT4o.	80
abstract_inverted_index.larger	67
abstract_inverted_index.models	69
abstract_inverted_index.recipe	119
abstract_inverted_index.single	88
abstract_inverted_index.strong	110
abstract_inverted_index.visual	31
abstract_inverted_index.Despite	54
abstract_inverted_index.Ovis-U1	25
abstract_inverted_index.achieve	123
abstract_inverted_index.between	97
abstract_inverted_index.compact	56
abstract_inverted_index.decoder	32
abstract_inverted_index.memory,	93
abstract_inverted_index.models.	134
abstract_inverted_index.operate	15
abstract_inverted_index.remains	84
abstract_inverted_index.results	105
abstract_inverted_index.systems	76
abstract_inverted_index.without	128
abstract_inverted_index.Seedream	78
abstract_inverted_index.achieves	59
abstract_inverted_index.backbone	112
abstract_inverted_index.combines	46
abstract_inverted_index.designed	13
abstract_inverted_index.high-end	89
abstract_inverted_index.indicate	106
abstract_inverted_index.moderate	92
abstract_inverted_index.pipeline	44
abstract_inverted_index.previous	24
abstract_inverted_index.reliable	124
abstract_inverted_index.stronger	35
abstract_inverted_index.tailored	51
abstract_inverted_index.training	43, 118
abstract_inverted_index.backbone,	39
abstract_inverted_index.bilingual	125
abstract_inverted_index.carefully	50, 115
abstract_inverted_index.combining	108
abstract_inverted_index.designed,	116
abstract_inverted_index.introduce	1
abstract_inverted_index.narrowing	94
abstract_inverted_index.optimized	8
abstract_inverted_index.oversized	131
abstract_inverted_index.practical	102
abstract_inverted_index.rendering	61, 100, 127
abstract_inverted_index.resorting	129
abstract_inverted_index.stringent	18
abstract_inverted_index.Crucially,	81
abstract_inverted_index.Ovis-Image	27, 58
abstract_inverted_index.Qwen-Image	72
abstract_inverted_index.approaches	74
abstract_inverted_index.deployable	85
abstract_inverted_index.framework,	26
abstract_inverted_index.integrates	28
abstract_inverted_index.leveraging	40
abstract_inverted_index.multimodal	38, 111
abstract_inverted_index.rendering,	12
abstract_inverted_index.sufficient	121
abstract_inverted_index.deployment.	103
abstract_inverted_index.efficiently	16
abstract_inverted_index.large-scale	47
abstract_inverted_index.performance	62
abstract_inverted_index.proprietary	133
abstract_inverted_index.constraints.	20
abstract_inverted_index.high-quality	10
abstract_inverted_index.pre-training	48
abstract_inverted_index.refinements.	53
abstract_inverted_index.specifically	7
abstract_inverted_index.text-centric	42
abstract_inverted_index.text-focused	117
abstract_inverted_index.architecture,	57
abstract_inverted_index.closed-source	75
abstract_inverted_index.computational	19
abstract_inverted_index.post-training	52
abstract_inverted_index.significantly	66
abstract_inverted_index.text-to-image	5
abstract_inverted_index.frontier-level	98
abstract_inverted_index.diffusion-based	30
abstract_inverted_index.$\textbf{Ovis-Image}$,	2
cited_by_percentile_year
countries_distinct_count	0
institutions_distinct_count	3
citation_normalized_percentile