CohereLanguage model
Command A Vision
Command A Vision has no overall index value yet because there are too few independent benchmark results.
PopularityRank 595 of 785
27out of 100-1 vs. last month
based on search, usage and attention
Summary
Command A Vision is Cohere's first Command A checkpoint for image inputs. The current model overview names charts, graphs, tables, OCR, document question answering, and object detection as enterprise tasks. It accepts text and images and returns text. Tool calls are not supported.
Cohere lists English, Portuguese, Italian, French, German, and Spanish support for this 112-billion-parameter checkpoint. Its context window holds 128,000 tokens. The vision variant adds image understanding to Command A, while tool calls remain with general Command A.
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 19 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.
ChartQA (Show measurement, test conditions, and source)
- Source value
- 90.9
- Score
- 90.9
- Metric
- score as published
- Unit
- %
- Benchmark version
- Command A Vision 07-2025; chart method not specified
- Category
- vision
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cohere Command A Vision release evaluation chart
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Command A Vision 07-2025 row in Cohere's evaluation figures; per-benchmark inference settings are not stated in the chart.
- Context
- Values are transcribed from a Cohere-published chart; the post notes that missing competitor reports may use best-effort internal evaluation.
InfoVQA (Show measurement, test conditions, and source)
- Source value
- 82.9
- Score
- 82.9
- Metric
- score as published
- Unit
- %
- Benchmark version
- Command A Vision 07-2025; chart method not specified
- Category
- vision
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cohere Command A Vision release evaluation chart
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Command A Vision 07-2025 row in Cohere's evaluation figures; per-benchmark inference settings are not stated in the chart.
- Context
- Values are transcribed from a Cohere-published chart; the post notes that missing competitor reports may use best-effort internal evaluation.
AI2D (Show measurement, test conditions, and source)
- Source value
- 94
- Score
- 94
- Metric
- score as published
- Unit
- %
- Benchmark version
- Command A Vision 07-2025; chart method not specified
- Category
- vision
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cohere Command A Vision release evaluation chart
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Command A Vision 07-2025 row in Cohere's evaluation figures; per-benchmark inference settings are not stated in the chart.
- Context
- Values are transcribed from a Cohere-published chart; the post notes that missing competitor reports may use best-effort internal evaluation.
MMMU (CoT) (Show measurement, test conditions, and source)
- Source value
- 65.3
- Score
- 65.3
- Metric
- chain-of-thought score as published
- Unit
- %
- Benchmark version
- MMMU; CoT; chart method not specified
- Category
- vision
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cohere Command A Vision release evaluation chart
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Command A Vision 07-2025 row in Cohere's evaluation figures; per-benchmark inference settings are not stated in the chart.
- Context
- Values are transcribed from a Cohere-published chart; the post notes that missing competitor reports may use best-effort internal evaluation.
MathVista (Show measurement, test conditions, and source)
- Source value
- 73.5
- Score
- 73.5
- Metric
- score as published
- Unit
- %
- Benchmark version
- MathVista; chart method not specified
- Category
- vision
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cohere Command A Vision release evaluation chart
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Command A Vision 07-2025 row in Cohere's evaluation figures; per-benchmark inference settings are not stated in the chart.
- Context
- Values are transcribed from a Cohere-published chart; the post notes that missing competitor reports may use best-effort internal evaluation.
DocVQA (Show measurement, test conditions, and source)
- Source value
- 95.9
- Score
- 95.9
- Metric
- score as published
- Unit
- %
- Benchmark version
- DocVQA; chart method not specified
- Category
- vision
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cohere Command A Vision release evaluation chart
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Command A Vision 07-2025 row in Cohere's evaluation figures; per-benchmark inference settings are not stated in the chart.
- Context
- Values are transcribed from a Cohere-published chart; the post notes that missing competitor reports may use best-effort internal evaluation.
TextVQA (Show measurement, test conditions, and source)
- Source value
- 84.8
- Score
- 84.8
- Metric
- score as published
- Unit
- %
- Benchmark version
- TextVQA; chart method not specified
- Category
- vision
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cohere Command A Vision release evaluation chart
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Command A Vision 07-2025 row in Cohere's evaluation figures; per-benchmark inference settings are not stated in the chart.
- Context
- Values are transcribed from a Cohere-published chart; the post notes that missing competitor reports may use best-effort internal evaluation.
OCRBench (Show measurement, test conditions, and source)
- Source value
- 86.9
- Score
- 86.9
- Metric
- score as published
- Unit
- %
- Benchmark version
- OCRBench; chart method not specified
- Category
- vision
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cohere Command A Vision release evaluation chart
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Command A Vision 07-2025 row in Cohere's evaluation figures; per-benchmark inference settings are not stated in the chart.
- Context
- Values are transcribed from a Cohere-published chart; the post notes that missing competitor reports may use best-effort internal evaluation.
RealWorldQA (Show measurement, test conditions, and source)
- Source value
- 73.6
- Score
- 73.6
- Metric
- score as published
- Unit
- %
- Benchmark version
- RealWorldQA; chart method not specified
- Category
- vision
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cohere Command A Vision release evaluation chart
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Command A Vision 07-2025 row in Cohere's evaluation figures; per-benchmark inference settings are not stated in the chart.
- Context
- Values are transcribed from a Cohere-published chart; the post notes that missing competitor reports may use best-effort internal evaluation.
CharXiv Reasoning (Show measurement, test conditions, and source)
- Source value
- 47
- Score
- 47
- Metric
- reported benchmark score on the reasoning track
- Unit
- %
- Benchmark version
- CharXiv
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cohere, Introducing Command A+, Figure 5 (Command A Vision comparison)
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact checkpoint command-a-vision-07-2025, identified by Cohere's official model documentation; Command A Vision is the exact model column in Figure 5 of Cohere's first-party article. The chart compares its CharXiv reasoning score with Command A+. Model serving and decoding settings are not reported.
- Context
- Cohere reports this in its own first-party article, so this is a vendor-reported observation rather than an independent reproduction. The article does not report a confidence interval; any unreported serving or decoding settings remain unknown.
CharXiv Descriptive (Show measurement, test conditions, and source)
- Source value
- 82
- Score
- 82
- Metric
- reported benchmark score on the descriptive track
- Unit
- %
- Benchmark version
- CharXiv
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Cohere, Introducing Command A+, Figure 5 (Command A Vision comparison)
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact checkpoint command-a-vision-07-2025, identified by Cohere's official model documentation; Command A Vision is the exact model column in Figure 5 of Cohere's first-party article. The chart compares its CharXiv descriptive score with Command A+. Model serving and decoding settings are not reported.
- Context
- Cohere reports this in its own first-party article, so this is a vendor-reported observation rather than an independent reproduction. The article does not report a confidence interval; any unreported serving or decoding settings remain unknown.
AECV-Bench object counting, door MAPE (Show measurement, test conditions, and source)
- Source value
- 69
- Score
- 69
- Metric
- Mean absolute percentage error (MAPE) for door counts
- Unit
- %
- Direction
- Lower is better
- Source type
- independent-evaluator
- Evaluator
- AECFoundry, AECV-Bench evaluators
- Status
- active
- Retrieved at
- 2026-10-23
- Methodology
- Independent evaluation. AECV-Bench object-counting evaluation on 120 floor plans. The paper lists Cohere Command A Vision and uses structured outputs with fixed fields for doors, windows, bedrooms, and toilets. Tables 1 and 2 report exact-match accuracy and mean absolute percentage error (MAPE), including a mean across the four classes.
- Context
- The study reports 120 floor plans as the evaluation-set size; this is not automatically the denominator of each MAPE cell. The per-cell MAPE denominator and handling of zero ground-truth counts are not specified. No confidence interval is reported.
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 101 models
As of November 2026 · Source: gradually.ai popularity index
Over time
Weekly values from October 18, 2026 to November 7, 2026
Command A VisionSolar Pro 2 28Command R7B Arabic 27Grok 4.5 27Claude Opus 4.7 27
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 Command A Vision, chosen by model line, developer, size, price and release date.
- Command A ReasoningCohere · same developer
- Command A TranslateCohere · same developer
- Command ACohere · same developer
- Command A PlusCohere · same developer
- North-Small-Translate-1.0Cohere · same developer
- Aya Vision 32BCohere · same developer
- Aya Expanse 32BCohere · same developer
- North Mini CodeCohere · same developer
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
Frequently asked questions about Command A Vision
What is the knowledge cutoff of Command A Vision?
June 2024, according to Cohere. That is 13 months before release.
Stated by Cohere
JunSepDecMarJunKnowledge cutoffJune 2024ReleasedJul 31, 2025
How large is the context window of Command A Vision?
128,000 tokens. That is how much text Command A Vision reads at once, about 190 A4 pages.
Page count as a rule of thumb from Artificial Analysis
GLM-5.1200K
DeepSeek-V3 0324164K
GLM-4.5131K
GPT OSS 120B131K
Command A VisionRank 98 of 103128K
DeepSeek-V3.2128K
DeepSeek-R1128K
Can I run Command A Vision locally?
Yes, the weights are open. With 112B parameters it needs a lot of memory, more a server than a laptop.
Stated by Cohere
- Open weights
- yes
- License
- CC BY-NC-4.0
- Parameters
- 112B
- active per token
- n/a
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.