Primary sources last checked on 13 Nov 2026. Verification freshness is tracked separately from factual change, and unknown values remain explicit rather than inferred.
NotableDistinct official first-party checkpoint with substantive specialist value; publisher rankings do not establish Leading.Curation method →
Open Model Weights records MiniMax-H3-Image as a Pending verification-parameter text to image and image editing open-weight model in the MiniMax H3 family, with MiniMax listed as the developer. The checked source repository provides 44 recognized weight artifacts, format evidence for Safetensors and a context window that is not declared in the checked standard sources. Values that are not evidenced in the checked sources remain explicitly undisclosed or unknown.
The repository declares other. Commercial-use classification requires license-specific review. Open Model Weights reports the checked license evidence and classification; this is not legal advice.
The checked sources declare MiniMaxAI/MiniMax-H3 as lineage evidence. Training recipe status: not-disclosed-in-checked-repository; data disclosure status: not-disclosed-in-standard-metadata-or-obvious-model-card-section.
MiniMax-H3-Image is an original downloadable text-to-image and image-editing model with a 33.1B diffusion transformer. Its November 10, 2026 community license excludes the EU, UK, South Korea and United States from the standard territory and adds commercial and hosted-service authorization conditions. Open Model Weights independently inspected the exact repository revision, listed weight artifacts and declared license; it has not run the model or reproduced publisher benchmark results. The repository includes a separately checked license file.
Source evidence 1The official H3-Image release describes a 33.1B dense diffusion transformer and native approximately 16-megapixel image generation; no OMW visual-quality or speed benchmarking has been performed.Source ↗
Source evidence 2The checkpoint relies on additional components including a VAE and H3 text encoder; a 68.65GB component download is not the complete deployment footprint.Source ↗
Source evidence 3The custom MiniMax-H3-Image Community License excludes the EU, UK, South Korea and United States from its applicable territory; ordinary access to public files is not authorization to use them there.Source ↗
Source evidence 4Section IV requires separate prior written permission above US$1M annualized revenue or for third-party hosted services, plus model attribution and further usage restrictions.Source ↗
Official immutable revision00dd081ebdd33315adf2c4d84f64aa327ddf3039Source ↗
Model license evidenceother; license-file-checkedSource ↗
Commercial-use statusunknown; see exact model-specific termsSource ↗
Repository cardREADME and source-described usage; no independent OMW inference runSource ↗
Performance provenancePublisher metrics, if mentioned, are not independent OMW measurementsSource ↗
WHY IT MATTERS
The publisher documents both text-to-image creation and image editing with multiple image references, using a dense diffusion transformer rather than a chat-oriented LLM.
The November 10 license defines the applicable territory as worldwide except the EU, UK, South Korea and United States. This is not equivalent to an unrestricted Apache or MIT release.
Statements about approximately 16MP output, six sampling steps and 4K-class image production come from MiniMax documentation and model cards, not OMW image tests.
MODEL-SELECTION VIEW
Strong fit when…
The publisher documents both text-to-image creation and image editing with multiple image references, using a dense diffusion transformer rather than a chat-oriented LLM.
The original weights and uncommon high-resolution editing architecture make the checkpoint a Notable technical release despite its restrictive usage rights.
The official H3-Image release describes a 33.1B dense diffusion transformer and native approximately 16-megapixel image generation; no OMW visual-quality or speed benchmarking has been performed.
Important considerations
The presence of public weights does not demonstrate that a workstation can run the complete stack: include encoder memory, VAE, inference buffers, data-transfer and batch-size overhead.
Separate written authorization is required above US$1M annualized revenue or for any third-party hosted service. Commercial interfaces require attribution, and the acceptable-use restrictions extend beyond ordinary copyright notices.
Review the license first, especially when operating in excluded jurisdictions. If authorized, plan GPU capacity for the diffusion core, text encoder, image decoder and expected image size.
Purpose, modality and checkpoint
The publisher documents both text-to-image creation and image editing with multiple image references, using a dense diffusion transformer rather than a chat-oriented LLM.
The claimed native output resolution and latency are first-party release specifications. Actual quality, composition, reference fidelity and speed require a declared hardware and prompt protocol.
Original weight artifacts and version identity
Inspect the original MiniMaxAI/MiniMax-H3-Image model tree, including transformer, VAE, tokenizer and auxiliary sampling files. The text encoder can be a separate dependency.
The presence of public weights does not demonstrate that a workstation can run the complete stack: include encoder memory, VAE, inference buffers, data-transfer and batch-size overhead.
License terms and commercial-use limits
The November 10 license defines the applicable territory as worldwide except the EU, UK, South Korea and United States. This is not equivalent to an unrestricted Apache or MIT release.
Separate written authorization is required above US$1M annualized revenue or for any third-party hosted service. Commercial interfaces require attribution, and the acceptable-use restrictions extend beyond ordinary copyright notices.
Publisher evaluations versus independent evidence
Statements about approximately 16MP output, six sampling steps and 4K-class image production come from MiniMax documentation and model cards, not OMW image tests.
Compare visual fidelity, edits, reproducibility, throughput and memory only with independent, reproducible procedures. Do not convert a publisher benchmark into OMW-measured performance.
Runtime integration and resource requirements
Official integrations mention diffusers and serving backends, but users must obtain each required component and verify compatible backend versions before assuming local viability.
Review the license first, especially when operating in excluded jurisdictions. If authorized, plan GPU capacity for the diffusion core, text encoder, image decoder and expected image size.
Practical selection and reproducibility
The original weights and uncommon high-resolution editing architecture make the checkpoint a Notable technical release despite its restrictive usage rights.
For readers in excluded territories, the principal practical answer is that public download does not mean local or commercial deployment is licensed. Seek authorization rather than assume permission.
IMPLEMENTATION & EVALUATION
How to evaluate MiniMax-H3-Image before deployment
Reproducible operator checklist; these steps are not claims of an OMW model run.
STEP 01
Pin original source and weight manifest
A named model does not guarantee identical checkpoint revisions.
REPRODUCIBLE ACCEPTANCE TEST
Record 40-character SHA, original file names, configuration and tokenizer/processor versions.
WHAT CAN GO WRONG
Unpinned revisions can silently change weights or usage requirements.
STEP 02
Reproduce task-specific inputs
A specialist checkpoint can fail outside the publisher's demonstrated conditions.
REPRODUCIBLE ACCEPTANCE TEST
Document actual prompt templates, modality preprocessing, inference backend and compatible software versions.
WHAT CAN GO WRONG
Using unsupported serving backends can invalidate comparisons.
STEP 03
Audit license and operational controls
Model card labels and code licenses are not automatically permission for every deployed use.
REPRODUCIBLE ACCEPTANCE TEST
Check exact weight terms, notices, tool permissions, data privacy and runtime isolation.
WHAT CAN GO WRONG
Restricted weight rights or unsafe action tools can make a deployment unsuitable.
STEP 04
Publish independent results only if measured
Vendor benchmark figures are useful evidence, but are not OMW-run observations.
REPRODUCIBLE ACCEPTANCE TEST
Pre-register held-out tasks and compare baseline checkpoint, hardware, quality, memory and error rates.
WHAT CAN GO WRONG
Attributing first-party results to independent inference would mislead readers.
The official H3-Image release describes a 33.1B dense diffusion transformer and native approximately 16-megapixel image generation; no OMW visual-quality or speed benchmarking has been performed.
Are the original MiniMax-H3-Image weights downloadable?
The official publisher repository is https://huggingface.co/MiniMaxAI/MiniMax-H3-Image. Download availability, original weight artifact names, and access rights are checked at the pinned source revision; download is not proof of deployment permission.
What license applies to MiniMax-H3-Image?
The custom MiniMax-H3-Image Community License excludes the EU, UK, South Korea and United States from its applicable territory; ordinary access to public files is not authorization to use them there.
Is MiniMax-H3-Image independently benchmarked by Open Model Weights?
No. The custom MiniMax-H3-Image Community License excludes the EU, UK, South Korea and United States from its applicable territory; ordinary access to public files is not authorization to use them there. Publisher tests and third-party claims are never presented as OMW-run inference or performance measurements.
Can MiniMax-H3-Image run locally?
Section IV requires separate prior written permission above US$1M annualized revenue or for third-party hosted services, plus model attribution and further usage restrictions.
Why is MiniMax-H3-Image classified Notable?
The official source-released weights and distinctive technical scope justify editorial attention. Notable is not a universal ranking; Leading would require a relevant independent, checkpoint-matched evaluation.
03 · PRACTICAL TECHNICAL PROFILE
Deployment, licensing and reproducible evidence.
Inspect weight-memory estimates, model lineage and runtime evidence separately from promotional or untested performance claims.
DEPLOYMENT PROFILE
Weight memory, without pretending it is a deployment guarantee.
These bars visualize weight-only decimal-GB estimates. They exclude KV cache, activations, runtime overhead, optimizer state and sharding.
WEIGHT-ONLY MEMORY
BF16 / FP16—
FP8 / INT8—
INT4—
FIELD EVIDENCE MAP
WeightsVerified
LicenseVerified
Commercial useUnknown
ContextNot declared
FormatsVerified
LineageDeclared
Training recipeNot disclosed in checked repository
Training dataNot disclosed in standard metadata or obvious model card section
Repository SHA 5f9047302397…; a source check is not an inference benchmark or legal clearance. See the per-model evidence JSON for the exact source record.
A verification check does not pretend the model changed. Open Model Weights updates the verification date after a real primary-source check, while factual-update dates move only when retained source evidence changes.
BF16: not observedFP16: not observedFP8: not observedINT8: not observedINT4: not observedGGUF: not observed
“Not observed” means no matching repository artifact/name/dtype signal was found; it does not prove no third-party conversion exists.
05 · LINEAGE
Base model & family
FamilyMiniMax H3
VariantH3 Image
Declared base modelMiniMaxAI/MiniMax-H3
Relationbase_model
06 · TRAINING ASSETS
Recipe & data disclosure
Training recipenot-disclosed-in-checked-repository
Data disclosurenot-disclosed-in-standard-metadata-or-obvious-model-card-section
EvaluationsNot classified by automated verifier
Intermediate checkpointsNot classified by automated verifier
07 · RUNTIME SUPPORT
Declared, mentioned and externally evidenced runtimes
Declared libraryminimax-h3
Declared/taggedminimax-h3
Mentioned in model cardComfyUI, Diffusers, SGLang, vLLM
Artifact signalsNone
External primary runtime evidenceNone
Tested by Open Model WeightsNo — source-derived support only
Open Model Weights separates model-repository evidence from external runtime-release evidence. An external support statement does not change the official model repository's weight formats and does not imply an Open Model Weights inference benchmark.
Official Hugging Face repository SHA unchanged; prior field evidence retained.
Field-by-field verification
Checked Hugging Face API metadata, exact repository file list, model card and config when accessible.
Field-by-field verification
Checked Hugging Face API metadata, exact repository file list, model card and config when accessible.
licenseverified
commercial useunknown
contextunknown
weight filesverified
formatsverified
base modeldeclared
training recipenot-disclosed-in-checked-repository
data disclosurenot-disclosed-in-standard-metadata-or-obvious-model-card-section
runtime supportsource-derived-not-runtime-tested
MODELS FOR AI AGENTS · ROLE EVIDENCE
Where this model fits in an agent system
Specialized model. A supporting component. This checkpoint does not, by category alone, serve as the planning and tool-execution controller.
The curated notable tier is not an agent-readiness rating. Model-specific tool-call success, long-horizon task accuracy and compatible runtime configurations are not independently established here.
Source-linked license: other (unknown commercial-use classification; see original terms).