MapikaDecision modelOpen source
Decider 2B Vision
PopularityRank 358 of 876
36out of 100-4 vs. last month
Summary
Decider 2B Vision adds images to the Decider readout. You provide an image, text, and typed questions, and the 1.9-billion-parameter checkpoint returns probabilities over your supplied options. Its context window holds 32,000 tokens.
For text-only inputs, Decider 2B is the counterpart. The vision variant adds an image to a text question, for example when inspecting a frame.
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 30 published benchmark results
Page 1 of 3
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.
Visual7W held-out (Show measurement, test conditions, and source)
- Source value
- 0.89
- Score
- 0.89
- Metric
- accuracy
- Unit
- ratio
- Category
- vision-decision
- Direction
- Higher is better
- Sample
- 300
- Source type
- vendor-reported
- Evaluator
- Mapika Decider vision model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. decider-2b-vision; Visual7W held-out; 300 items per task.
A-OKVQA (Show measurement, test conditions, and source)
- Source value
- 0.85
- Score
- 0.85
- Metric
- accuracy
- Unit
- ratio
- Category
- vision-decision
- Direction
- Higher is better
- Sample
- 300
- Source type
- vendor-reported
- Evaluator
- Mapika Decider vision model card
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. decider-2b-vision; A-OKVQA; 300 items per task.
Pong RAM-state teacher agreement (Show measurement, test conditions, and source)
- Source value
- 0.96
- Score
- 0.96
- Metric
- agreement with RAM-state teacher
- Unit
- ratio
- Benchmark version
- Mapika/decider MODEL_CARD_VISION.md; repository revision d6fc522aed7fc392576a9b66a330409bc6c382bd; text weights v5
- Direction
- Higher is better
- Sample
- 300
- Source type
- vendor-reported
- Evaluator
- Mapika, Decider vision model card
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. Decider-2B-Vision v5 text weights transplanted into Qwen3.5-2B; the official card reports 300 cases per task; card revision d6fc522aed7fc392576a9b66a330409bc6c382bd.
- Context
- These are exact card figures for its v5-derived vision experiment, not general held-out visual reasoning. Some task material or supervision is in the documented training sources.
Breakout RAM-state teacher agreement (Show measurement, test conditions, and source)
- Source value
- 0.96
- Score
- 0.96
- Metric
- agreement with RAM-state teacher
- Unit
- ratio
- Benchmark version
- Mapika/decider MODEL_CARD_VISION.md; repository revision d6fc522aed7fc392576a9b66a330409bc6c382bd; text weights v5
- Direction
- Higher is better
- Sample
- 300
- Source type
- vendor-reported
- Evaluator
- Mapika, Decider vision model card
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. Decider-2B-Vision v5 text weights transplanted into Qwen3.5-2B; the official card reports 300 cases per task; card revision d6fc522aed7fc392576a9b66a330409bc6c382bd.
- Context
- These are exact card figures for its v5-derived vision experiment, not general held-out visual reasoning. Some task material or supervision is in the documented training sources.
AI2D (Show measurement, test conditions, and source)
- Source value
- 0.93
- Score
- 0.93
- Metric
- accuracy
- Unit
- ratio
- Benchmark version
- Mapika/decider MODEL_CARD_VISION.md; repository revision d6fc522aed7fc392576a9b66a330409bc6c382bd; text weights v5
- Direction
- Higher is better
- Sample
- 300
- Source type
- vendor-reported
- Evaluator
- Mapika, Decider vision model card
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. Decider-2B-Vision v5 text weights transplanted into Qwen3.5-2B; the official card reports 300 cases per task; card revision d6fc522aed7fc392576a9b66a330409bc6c382bd.
- Context
- These are exact card figures for its v5-derived vision experiment, not general held-out visual reasoning. Some task material or supervision is in the documented training sources.
ScienceQA (Show measurement, test conditions, and source)
- Source value
- 0.95
- Score
- 0.95
- Metric
- accuracy
- Unit
- ratio
- Benchmark version
- Mapika/decider MODEL_CARD_VISION.md; repository revision d6fc522aed7fc392576a9b66a330409bc6c382bd; text weights v5
- Direction
- Higher is better
- Sample
- 300
- Source type
- vendor-reported
- Evaluator
- Mapika, Decider vision model card
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. Decider-2B-Vision v5 text weights transplanted into Qwen3.5-2B; the official card reports 300 cases per task; card revision d6fc522aed7fc392576a9b66a330409bc6c382bd.
- Context
- These are exact card figures for its v5-derived vision experiment, not general held-out visual reasoning. Some task material or supervision is in the documented training sources.
IconQA (Show measurement, test conditions, and source)
- Source value
- 0.94
- Score
- 0.94
- Metric
- accuracy
- Unit
- ratio
- Benchmark version
- Mapika/decider MODEL_CARD_VISION.md; repository revision d6fc522aed7fc392576a9b66a330409bc6c382bd; text weights v5
- Direction
- Higher is better
- Sample
- 300
- Source type
- vendor-reported
- Evaluator
- Mapika, Decider vision model card
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. Decider-2B-Vision v5 text weights transplanted into Qwen3.5-2B; the official card reports 300 cases per task; card revision d6fc522aed7fc392576a9b66a330409bc6c382bd.
- Context
- These are exact card figures for its v5-derived vision experiment, not general held-out visual reasoning. Some task material or supervision is in the documented training sources.
Raven (Show measurement, test conditions, and source)
- Source value
- 0.8
- Score
- 0.8
- Metric
- accuracy
- Unit
- ratio
- Benchmark version
- Mapika/decider MODEL_CARD_VISION.md; repository revision d6fc522aed7fc392576a9b66a330409bc6c382bd; text weights v5
- Direction
- Higher is better
- Sample
- 300
- Source type
- vendor-reported
- Evaluator
- Mapika, Decider vision model card
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. Decider-2B-Vision v5 text weights transplanted into Qwen3.5-2B; the official card reports 300 cases per task; card revision d6fc522aed7fc392576a9b66a330409bc6c382bd.
- Context
- These are exact card figures for its v5-derived vision experiment, not general held-out visual reasoning. Some task material or supervision is in the documented training sources.
Hateful Memes (Show measurement, test conditions, and source)
- Source value
- 0.8
- Score
- 0.8
- Metric
- accuracy
- Unit
- ratio
- Benchmark version
- Mapika/decider MODEL_CARD_VISION.md; repository revision d6fc522aed7fc392576a9b66a330409bc6c382bd; text weights v5
- Direction
- Higher is better
- Sample
- 300
- Source type
- vendor-reported
- Evaluator
- Mapika, Decider vision model card
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. Decider-2B-Vision v5 text weights transplanted into Qwen3.5-2B; the official card reports 300 cases per task; card revision d6fc522aed7fc392576a9b66a330409bc6c382bd.
- Context
- These are exact card figures for its v5-derived vision experiment, not general held-out visual reasoning. Some task material or supervision is in the documented training sources.
NaturalBench overall accuracy (Show measurement, test conditions, and source)
- Source value
- 0.7942
- Score
- 0.7942
- Metric
- accuracy
- Unit
- ratio
- Benchmark version
- NaturalBench dataset revision ba41a7d564877a9b64c094b08015ca493cc3e54b
- Category
- vision
- Direction
- Higher is better
- Sample
- 1200
- Source type
- community-reported
- Evaluator
- Strands Decider vision evaluation, GitHub release asset vision-eval-2026-10-03
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Community-reported result. Mapika/decider-2b-vision@863e290863655f1d6b69324d77d09ac972d21609, bfloat16; evaluation script at commit 333b7f4a0662f8e8bd6d8eb9b193c24a75c9f032, torch 2.7.1+cu126, transformers 5.18.0, 1x NVIDIA H100 80GB. Mapika's own visual model code and prompt, one image resized to at most 768x768 pixels. NaturalBench, 300 groups x 2 images x 2 questions; one item is one image-question pair. Scope: overall.
- Context
- Community-published evaluation, not an independently verified run. It uses Mapika's own model loader and answer-slot prompt. The report pins the evaluated model and datasets but does not provide an independently reproduced result.
NaturalBench overall multiclass Brier score (Show measurement, test conditions, and source)
- Source value
- 0.3036
- Score
- 0.3036
- Metric
- multiclass Brier score
- Unit
- ratio
- Benchmark version
- NaturalBench dataset revision ba41a7d564877a9b64c094b08015ca493cc3e54b
- Category
- vision
- Direction
- Lower is better
- Sample
- 1200
- Source type
- community-reported
- Evaluator
- Strands Decider vision evaluation, GitHub release asset vision-eval-2026-10-03
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Community-reported result. Mapika/decider-2b-vision@863e290863655f1d6b69324d77d09ac972d21609, bfloat16; evaluation script at commit 333b7f4a0662f8e8bd6d8eb9b193c24a75c9f032, torch 2.7.1+cu126, transformers 5.18.0, 1x NVIDIA H100 80GB. Mapika's own visual model code and prompt, one image resized to at most 768x768 pixels. NaturalBench, 300 groups x 2 images x 2 questions; one item is one image-question pair. Scope: overall.
- Context
- Community-published evaluation, not an independently verified run. It uses Mapika's own model loader and answer-slot prompt. The report pins the evaluated model and datasets but does not provide an independently reproduced result.
NaturalBench overall expected calibration error (10 equal-width bins) (Show measurement, test conditions, and source)
- Source value
- 0.0799
- Score
- 0.0799
- Metric
- expected calibration error (10 equal-width bins)
- Unit
- ratio
- Benchmark version
- NaturalBench dataset revision ba41a7d564877a9b64c094b08015ca493cc3e54b
- Category
- vision
- Direction
- Lower is better
- Sample
- 1200
- Source type
- community-reported
- Evaluator
- Strands Decider vision evaluation, GitHub release asset vision-eval-2026-10-03
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Community-reported result. Mapika/decider-2b-vision@863e290863655f1d6b69324d77d09ac972d21609, bfloat16; evaluation script at commit 333b7f4a0662f8e8bd6d8eb9b193c24a75c9f032, torch 2.7.1+cu126, transformers 5.18.0, 1x NVIDIA H100 80GB. Mapika's own visual model code and prompt, one image resized to at most 768x768 pixels. NaturalBench, 300 groups x 2 images x 2 questions; one item is one image-question pair. Scope: overall.
- Context
- Community-published evaluation, not an independently verified run. It uses Mapika's own model loader and answer-slot prompt. The report pins the evaluated model and datasets but does not provide an independently reproduced result.
Page 1 of 3
Popularity compared
How often the model is searched for, used and mentioned, compared with every model of its category.
Popularity index
0 to 100, higher is more popular
20 of 37 models
As of November 2026 · Source: gradually.ai popularity index
Over time
Weekly values from October 18, 2026 to November 6, 2026
Decider 2B VisionIntern-Decision-2B 37CLM-v0.1-8B 37Jeff 1 36GLiNER2.5-Decide-1B 34
Method changes in the timeline, a jump at these points is no market event:
- 1November 1, 2026: Value only with search or usage data, 30-day attention
What is gaining popularity right now, across models, vendors, harnesses and AI subscriptions, is collected in All AI trends.
Similar models
More models of the same kind as Decider 2B Vision, chosen by model line, developer, size, price and release date.
All specifications
Architecture
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
Sources and data date
Show all 17 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.