InclusionAIImage model
Ming-Image-0.1-Design
Arena ratingnot rated
n/aUser voteshigher is better
GraduallyBenchnot tested yet
n/aour own rating0 to 100
Priceno price
n/aper imagelower is better
PopularityRank 385 of 875
34out of 100same as last month
Summary
Ming-Image-0.1-Design is a 6-billion-parameter model for interfaces, infographics, posters, and other text-rich designs. It produces complete compositions, returns RGBA images with transparent backgrounds, and recommends 2048 x 2048, with 1024 x 1024 for faster runs.
The open MIT weights are published on Hugging Face and ModelScope. The inference setup validates one CUDA GPU with 80 GiB of VRAM.
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 14 published benchmark results
Page 1 of 2
Measurements without matching peer values
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
ImageBench capability pass rate (Show measurement, test conditions, and source)
- Source value
- 61.5
- Score
- 61.5
- Metric
- Capability pass rate (0-100)
- Unit
- %
- Benchmark version
- ImageBench V1 exact model-page observation; run date not stated
- Category
- text-to-image
- Direction
- Higher is better
- Sample
- 192
- Source type
- independent-evaluator
- Evaluator
- ImageBench V1 (exact model page)
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. ImageBench row label local/ming-image-0.1-design-6b; run target gx10/ming-image-0.1-design-6b; 1024x1024, 12 inference steps, seed 7, BasicGuider cfg 1.0; 192 generated images.
- Context
- 62% is the page's integer-rounded display. The underlying counts imply a more precise arithmetic fraction, but that derived precision is not substituted for the published value. This is a V1 page result, not an exact row from the pinned V1.2 scores.csv snapshot. ImageBench does not publish a weight hash, runtime precision, or quantization for the gx10 target; byte identity to the vendor files is unverified.
ImageBench category pass rate (Show measurement, test conditions, and source)
- Source value
- 66.7
- Score
- 66.7
- Metric
- Graphical design pass rate
- Unit
- %
- Benchmark version
- ImageBench V1 exact model-page observation; run date not stated
- Category
- text-to-image
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- ImageBench V1 (exact model page)
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. ImageBench row label local/ming-image-0.1-design-6b, text-to-image; 1024x1024, 12 inference steps, seed 7, BasicGuider cfg 1.0. The embedded category summary does not publish a category-specific denominator.
- Context
- The page displays category rates as whole percentages, without category-specific denominators. Do not derive N from 192 total generations. This is an ImageBench V1 page value; no exact V1.2 dataset row was found. ImageBench does not publish a weight hash, runtime precision, or quantization for the gx10 target; byte identity to the vendor files is unverified.
ImageBench category pass rate (Show measurement, test conditions, and source)
- Source value
- 54.8
- Score
- 54.8
- Metric
- Human realism pass rate
- Unit
- %
- Benchmark version
- ImageBench V1 exact model-page observation; run date not stated
- Category
- text-to-image
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- ImageBench V1 (exact model page)
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. ImageBench row label local/ming-image-0.1-design-6b, text-to-image; 1024x1024, 12 inference steps, seed 7, BasicGuider cfg 1.0. The embedded category summary does not publish a category-specific denominator.
- Context
- The page displays category rates as whole percentages, without category-specific denominators. Do not derive N from 192 total generations. This is an ImageBench V1 page value; no exact V1.2 dataset row was found. ImageBench does not publish a weight hash, runtime precision, or quantization for the gx10 target; byte identity to the vendor files is unverified.
ImageBench category pass rate (Show measurement, test conditions, and source)
- Source value
- 85.2
- Score
- 85.2
- Metric
- Professional Studio pass rate
- Unit
- %
- Benchmark version
- ImageBench V1 exact model-page observation; run date not stated
- Category
- text-to-image
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- ImageBench V1 (exact model page)
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. ImageBench row label local/ming-image-0.1-design-6b, text-to-image; 1024x1024, 12 inference steps, seed 7, BasicGuider cfg 1.0. The embedded category summary does not publish a category-specific denominator.
- Context
- The page displays category rates as whole percentages, without category-specific denominators. Do not derive N from 192 total generations. This is an ImageBench V1 page value; no exact V1.2 dataset row was found. ImageBench does not publish a weight hash, runtime precision, or quantization for the gx10 target; byte identity to the vendor files is unverified.
ImageBench category pass rate (Show measurement, test conditions, and source)
- Source value
- 61.4
- Score
- 61.4
- Metric
- Spatial Reasoning pass rate
- Unit
- %
- Benchmark version
- ImageBench V1 exact model-page observation; run date not stated
- Category
- text-to-image
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- ImageBench V1 (exact model page)
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. ImageBench row label local/ming-image-0.1-design-6b, text-to-image; 1024x1024, 12 inference steps, seed 7, BasicGuider cfg 1.0. The embedded category summary does not publish a category-specific denominator.
- Context
- The page displays category rates as whole percentages, without category-specific denominators. Do not derive N from 192 total generations. This is an ImageBench V1 page value; no exact V1.2 dataset row was found. ImageBench does not publish a weight hash, runtime precision, or quantization for the gx10 target; byte identity to the vendor files is unverified.
ImageBench category pass rate (Show measurement, test conditions, and source)
- Source value
- 66.7
- Score
- 66.7
- Metric
- Text Rendering pass rate
- Unit
- %
- Benchmark version
- ImageBench V1 exact model-page observation; run date not stated
- Category
- text-to-image
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- ImageBench V1 (exact model page)
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. ImageBench row label local/ming-image-0.1-design-6b, text-to-image; 1024x1024, 12 inference steps, seed 7, BasicGuider cfg 1.0. The embedded category summary does not publish a category-specific denominator.
- Context
- The page displays category rates as whole percentages, without category-specific denominators. Do not derive N from 192 total generations. This is an ImageBench V1 page value; no exact V1.2 dataset row was found. ImageBench does not publish a weight hash, runtime precision, or quantization for the gx10 target; byte identity to the vendor files is unverified.
ImageBench category pass rate (Show measurement, test conditions, and source)
- Source value
- 40.7
- Score
- 40.7
- Metric
- Truthfulness pass rate
- Unit
- %
- Benchmark version
- ImageBench V1 exact model-page observation; run date not stated
- Category
- text-to-image
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- ImageBench V1 (exact model page)
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. ImageBench row label local/ming-image-0.1-design-6b, text-to-image; 1024x1024, 12 inference steps, seed 7, BasicGuider cfg 1.0. The embedded category summary does not publish a category-specific denominator.
- Context
- The page displays category rates as whole percentages, without category-specific denominators. Do not derive N from 192 total generations. This is an ImageBench V1 page value; no exact V1.2 dataset row was found. ImageBench does not publish a weight hash, runtime precision, or quantization for the gx10 target; byte identity to the vendor files is unverified.
AA Image T2I (Show measurement, test conditions, and source)
- Source value
- 998
- Score
- 998
- Metric
- elo
- Unit
- elo
- Benchmark version
- AA Image T2I v2.0
- Category
- image-quality
- Direction
- Higher is better
- Sample
- 21566
- Confidence interval
- 8
- Lower confidence bound
- 990
- Upper confidence bound
- 1006
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Independent evaluation. Text-to-image pairwise human-preference Arena; exact Ming-Image-0.1-Design row; rank 44 with range 40-52.
- Context
- Independent Arena evidence. An undated vendor UI/UX screenshot showing 1,082 Elo is excluded because it has no dated cohort, sample count, or interval.
AA Image Arena Elo (Show measurement, test conditions, and source)
- Source value
- 998
- Score
- 998
- Metric
- pairwise human-preference Elo
- Unit
- elo
- Benchmark version
- AA-Image-T2I v2.1
- Category
- text-to-image
- Direction
- Higher is better
- Sample
- 19976
- Confidence interval
- 9
- Lower confidence bound
- 989
- Upper confidence bound
- 1007
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis AA-Image-T2I v2.1
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Overall text-to-image; exact leaderboard row Ming-Image-0.1-Design.
Vendor-republished UI/UX Design Elo (Show measurement, test conditions, and source)
- Source value
- 1082
- Score
- 1082
- Metric
- Elo rating from blind image-preference votes
- Unit
- elo
- Category
- ui-ux-design
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- InclusionAI official model card (Artificial Analysis Image Arena screenshot)
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. Text-to-image UI/UX Design category; exact screenshot row Ming-Image-0.1-Design on the Open Weights leaderboard. Prompt cohort, settings and sample population are not stated.
- Context
- The evaluation date is not disclosed. This is a vendor-republished screenshot, not a live evaluator readback. It publishes no numeric sample count or confidence interval; the screenshot's bar and whisker were not used to infer either. The score is category-specific and must not be treated as the model's overall text-to-image score or as directly comparable to the current AA overall cohort.
ImageBench (Show measurement, test conditions, and source)
- Source value
- 64.9
- Score
- 64.9
- Metric
- Overall composite (capability and estimated preference; 0-100 scale)
- Unit
- points
- Benchmark version
- ImageBench V1 exact model-page observation; run date not stated
- Category
- text-to-image
- Direction
- Higher is better
- Sample
- 192
- Source type
- independent-evaluator
- Evaluator
- ImageBench V1 (exact model page)
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. ImageBench row label local/ming-image-0.1-design-6b; the inclusionAI model card separately confirms the Ming-Image-0.1-Design 6B name, not gx10 checkpoint bytes. Model page gives run target gx10/ming-image-0.1-design-6b, 1024x1024, 12 inference steps, seed 7, unguided BasicGuider cfg 1.0; 192 generations.
- Context
- The model-specific page is labeled ImageBench V1 and its run date and judge revision are not stated. A pinned V1.2 scores.csv snapshot inspected in this round contains 50 model rows but no exact Ming row. This is retained as the page's V1 observation, not asserted as a V1.2 result. ImageBench warns that VLM verdicts may be wrong and its prompt set favors separating models. ImageBench does not publish a weight hash, runtime precision, or quantization for the gx10 target; byte identity to the vendor files is unverified.
ImageBench (Show measurement, test conditions, and source)
- Source value
- 68.3
- Score
- 68.3
- Metric
- Estimated Preference Score (HPSv3-based; 0-100)
- Unit
- points
- Benchmark version
- ImageBench V1 exact model-page observation; run date not stated
- Category
- text-to-image
- Direction
- Higher is better
- Sample
- 165
- Source type
- independent-evaluator
- Evaluator
- ImageBench V1 (exact model page)
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. The model page embedded result payload publishes eps=68.3 and eps_n=165. The publisher does not explain how this EPS denominator is constructed; it is distinct from the 192 generated images.
- Context
- The page publishes eps_n=165, but the methodology for constructing that denominator is not stated. It is not the total of 192 generated images. The page is ImageBench V1; the inspected V1.2 CSV has no exact Ming row.
Page 1 of 2
Popularity compared
How often the model is searched for, used and mentioned, compared with every model of its category.
Popularity index
The 100 most popular models of this category, 0 to 100, higher is more popular
20 of 100 models
Ming-Image-0.1-Design34
Stable Diffusion XL 1.0, Stability AI: 72
Anima Base v1.0, CircleStone Labs and Comfy Org: 68
SDXL Turbo, Stability AI: 63
Stable Diffusion Turbo, Stability AI: 63
Nano Banana 2, Google: 61
FLUX.2 Max, Black Forest Labs: 55
Krea 2 Raw, Krea.ai: 54
Nano Banana Pro, Google: 50
FLUX.2 Pro, Black Forest Labs: 48
MOSS-Image-SFT, OpenMOSS: 46
FLUX.2 Flex, Black Forest Labs: 46
GPT Image 2, OpenAI: 46
LongCat-Image, Meituan: 45
MAI-Image-2.6, Microsoft AI: 45
FLUX.2 Klein 9B, Black Forest Labs: 45
PixArt-Sigma-XL-2-1024-MS, PixArt-Alpha: 44
Janus-Pro-7B, DeepSeek: 44
HunyuanImage 3.0 Instruct, Tencent: 44
MOSS-Image-Base, OpenMOSS: 44
Ming-Image-0.1-Design, InclusionAI: 34
Ming-Image-0.1-Design
As of November 2026 · Source: gradually.ai popularity index
Over time
Weekly values from October 18, 2026 to November 5, 2026
Ming-Image-0.1-DesignQwen Image 2.0 ProMAI-Image-2.5 FlashSeedream 5.0 ProMidjourney V8.1
Ming-Image-0.1-Design compared
Compare with another model
All specifications
Architecture
- Checkpoint
- Independent 6B text-to-image checkpoint; do not confuse it with Ming Design Layer.
- Source: Checkpoint
Input and output
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
API
Further details
- Hosted route
- DeepInfra hosts the checkpoint as inclusionAI/Ming-Image-0.1-Design.
- Source: Hosted route
- Tasks
- Generates complete images from text; the Layer checkpoint instead requires an input image.
- Source: Tasks
Sources and data date
Show all 14 sources
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
No active result from the reviewed external evaluators can currently be attributed to this model. Other source evidence remains available below.
Without vendor claims and historical or disputed results; those stay labelled in the single values.
| Type | Evidence and data date |
|---|---|
| Research status | Research date October 4, 2026. 8 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Not found in the inspected sources | ming-image-0-1-design-aa-editing-row Note The current AA editing leaderboard has no exact Ming-Image-0.1-Design row. Its text-to-image result and layer-decomposition scores are not image-editing results and are not transferred. |
| Not found in the inspected sources | imagebench-v1-2-exact-ming-row Note The captured, pinned V1.2 scores.csv has 50 model rows and no exact local--ming-image-0.1-design-6b row. The separate exact model page is labeled V1 and remains a separate record. |
| Unresolved | ming-imagebench-run-date-and-judge-revision Note The exact model page does not publish the run date or VLM judge revision; neither is inferred from the capture date. |
| Additional source | InclusionAI (retrieved October 3, 2026; October 16, 2026; November 1, 2026) · Ming-Image Design model card · Editorial description reviewed October 3, 2026 |
| Additional source | InclusionAI (retrieved November 1, 2026) |
| Additional source | InclusionAI (retrieved October 3, 2026; November 1, 2026) · Ming-Image repository · Editorial description reviewed October 3, 2026 |
| Additional source | Artificial Analysis (retrieved October 4, 2026; October 16, 2026; October 19, 2026) |
| Additional source | ImageBench V1 (exact model page) (retrieved October 16, 2026) |
| Additional source | Model metadata (retrieved October 16, 2026)123 |
| Additional source | OpenRouter: Popularity index (derived), CC BY 4.0 |
| Additional source | Vercel AI Gateway: Popularity index (derived), CC BY 4.0 |
| Additional source | Ollama: cumulative downloads for the popularity index |
| Additional source | Replicate: cumulative runs for the popularity index |
| Additional source | Google Trending Now: limited feed selection for Germany and the US |
| Additional source | Creative Commons Attribution 4.0 |