CohereLanguage model
Aya Expanse 8B
Aya Expanse 8B has no overall index value yet because there are too few independent benchmark results.
PopularityRank 280 of 785
40out of 100+1 vs. last month
based on search, usage and attention
Summary
Aya Expanse 8B brings Cohere's 23-language Aya approach into an open 8-billion-parameter checkpoint. The underlying research combines multilingual preference training, safety tuning, and model merging. The checkpoint consequently stays focused on translation and instruction tasks, without visual understanding or built-in tools.
The 8-billion-parameter model provides an 8,000-token context, considerably narrower than Aya Expanse 32B. The weights use CC-BY-NC-4.0, which rules out unrestricted commercial use.
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 58 published benchmark results
Page 1 of 5
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.
Discriminative average (Show measurement, test conditions, and source)
- Source value
- 70.3
- Score
- 70.3
- Metric
- average accuracy over XWinograd, XCOPA and XStoryCloze
- Unit
- %
- Benchmark version
- Aya Expanse paper; 0-shot; 23-language evaluation
- Category
- multilingual
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Aya Expanse authors' multilingual evaluation paper
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Aya Expanse 8B results. Global-MMLU is 5-shot; discriminative tasks are zero-shot; MGSM is 5-shot CoT/strict-match; translation uses FLORES-200 and xCOMET-XL.
- Context
- The source is the model authors' paper. It does not disclose every generation detail or evaluation date; no later Aya checkpoint is implied.
Global-MMLU (Show measurement, test conditions, and source)
- Source value
- 53.7
- Score
- 53.7
- Metric
- accuracy; 5-shot
- Unit
- %
- Benchmark version
- Aya Expanse paper; 23-language multilingual MMLU
- Category
- multilingual
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Aya Expanse authors' multilingual evaluation paper
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Aya Expanse 8B results. Global-MMLU is 5-shot; discriminative tasks are zero-shot; MGSM is 5-shot CoT/strict-match; translation uses FLORES-200 and xCOMET-XL.
- Context
- The source is the model authors' paper. It does not disclose every generation detail or evaluation date; no later Aya checkpoint is implied.
MGSM (Show measurement, test conditions, and source)
- Source value
- 67
- Score
- 67
- Metric
- strict-match accuracy; 5-shot chain-of-thought
- Unit
- %
- Benchmark version
- Aya Expanse paper; 7 languages
- Category
- multilingual
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Aya Expanse authors' multilingual evaluation paper
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Vendor-reported result. Exact Aya Expanse 8B results. Global-MMLU is 5-shot; discriminative tasks are zero-shot; MGSM is 5-shot CoT/strict-match; translation uses FLORES-200 and xCOMET-XL.
- Context
- The source is the model authors' paper. It does not disclose every generation detail or evaluation date; no later Aya checkpoint is implied.
INCLUDE accuracy (Show measurement, test conditions, and source)
- Source value
- 37.2
- Score
- 37.2
- Metric
- accuracy
- Unit
- %
- Category
- multilingual-understanding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Aya Expanse technical report, arXiv:2412.04261v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. INCLUDE-base, zero-shot native-language chain-of-thought, 44 languages.
m-ArenaHard win rate vs Gemma 2 9B (Show measurement, test conditions, and source)
- Source value
- 60.4
- Score
- 60.4
- Metric
- pairwise win rate
- Unit
- %
- Category
- open-ended-generation
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Aya Expanse technical report, arXiv:2412.04261v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. m-ArenaHard pairwise comparison, averaged across 23 languages; GPT-4o-2024-08-06 judge; instruct-tuned competitors.
m-ArenaHard win rate vs Qwen 2.5 7B (Show measurement, test conditions, and source)
- Source value
- 55.7
- Score
- 55.7
- Metric
- pairwise win rate
- Unit
- %
- Category
- open-ended-generation
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Aya Expanse technical report, arXiv:2412.04261v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. m-ArenaHard pairwise comparison, averaged across 23 languages; GPT-4o-2024-08-06 judge; instruct-tuned competitors.
m-ArenaHard win rate vs Llama 3.1 8B (Show measurement, test conditions, and source)
- Source value
- 70.6
- Score
- 70.6
- Metric
- pairwise win rate
- Unit
- %
- Category
- open-ended-generation
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Aya Expanse technical report, arXiv:2412.04261v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. m-ArenaHard pairwise comparison, averaged across 23 languages; GPT-4o-2024-08-06 judge; instruct-tuned competitors.
Dolly win rate vs Llama 3.1 8B (Show measurement, test conditions, and source)
- Source value
- 83.9
- Score
- 83.9
- Metric
- pairwise win rate
- Unit
- %
- Category
- open-ended-generation
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Aya Expanse technical report, arXiv:2412.04261v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Dolly held-out evaluation, averaged across 23 languages; GPT-4o-2024-08-06 judge; instruct-tuned competitors.
- Context
- The authors describe Dolly prompts as less difficult and note distribution overlap with Dolly training data.
ALARB verdict prediction / Facts Only / correct-verdict rate (Show measurement, test conditions, and source)
- Source value
- 25.6
- Score
- 25.6
- Metric
- Correct-verdict rate
- Unit
- %
- Benchmark version
- ALARB (ArabicNLP 2025), Table 4
- Category
- legal-reasoning
- Direction
- Higher is better
- Sample
- 1329
- Source type
- independent-evaluator
- Evaluator
- ALARB: An Arabic Legal Argument Reasoning Benchmark (ArabicNLP 2025; DOI 10.18653/v1/2025.arabicnlp-main.32)
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Independent evaluation. aya-expanse-8B, exact model row in Table 4; verdict prediction with only the case facts, without the court's reasoning; the model is asked to reason and predict a verdict; GPT-4o LLM-as-a-judge classifies alignment with the actual verdict; Table 4 correct-verdict rate for Facts Only.
- Context
- Section 5 states that the evaluation of all tasks used a subset of 1,329 legal cases; this reported evaluation set is applied as N for each Table 4 verdict-prediction task. GPT-4o was used as an LLM judge. No API/checkpoint revision, decoding configuration, or confidence interval is published. Partial and incorrect outcome shares are categorical diagnostics, not separate monotonic quality scores, and are not imported.
ALARB verdict prediction / Facts & Reasoning / correct-verdict rate (Show measurement, test conditions, and source)
- Source value
- 61.9
- Score
- 61.9
- Metric
- Correct-verdict rate
- Unit
- %
- Benchmark version
- ALARB (ArabicNLP 2025), Table 4
- Category
- legal-reasoning
- Direction
- Higher is better
- Sample
- 1329
- Source type
- independent-evaluator
- Evaluator
- ALARB: An Arabic Legal Argument Reasoning Benchmark (ArabicNLP 2025; DOI 10.18653/v1/2025.arabicnlp-main.32)
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Independent evaluation. aya-expanse-8B, exact model row in Table 4; verdict prediction with case facts plus the court's reasoning; predict the final verdict; GPT-4o LLM-as-a-judge classifies alignment with the actual verdict; Table 4 correct-verdict rate for Facts & Reasoning.
- Context
- Section 5 states that the evaluation of all tasks used a subset of 1,329 legal cases; this reported evaluation set is applied as N for each Table 4 verdict-prediction task. GPT-4o was used as an LLM judge. No API/checkpoint revision, decoding configuration, or confidence interval is published. Partial and incorrect outcome shares are categorical diagnostics, not separate monotonic quality scores, and are not imported.
ALARB verdict prediction / Facts & Regulations / correct-verdict rate (Show measurement, test conditions, and source)
- Source value
- 24.6
- Score
- 24.6
- Metric
- Correct-verdict rate
- Unit
- %
- Benchmark version
- ALARB (ArabicNLP 2025), Table 4
- Category
- legal-reasoning
- Direction
- Higher is better
- Sample
- 1329
- Source type
- independent-evaluator
- Evaluator
- ALARB: An Arabic Legal Argument Reasoning Benchmark (ArabicNLP 2025; DOI 10.18653/v1/2025.arabicnlp-main.32)
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Independent evaluation. aya-expanse-8B, exact model row in Table 4; verdict prediction with case facts plus relevant legal articles; reason and predict a verdict; GPT-4o LLM-as-a-judge classifies alignment with the actual verdict; Table 4 correct-verdict rate for Facts & Regulations.
- Context
- Section 5 states that the evaluation of all tasks used a subset of 1,329 legal cases; this reported evaluation set is applied as N for each Table 4 verdict-prediction task. GPT-4o was used as an LLM judge. No API/checkpoint revision, decoding configuration, or confidence interval is published. Partial and incorrect outcome shares are categorical diagnostics, not separate monotonic quality scores, and are not imported.
Macaron Table 8, logprob, MC-EN (Show measurement, test conditions, and source)
- Source value
- 52.67
- Score
- 52.67
- Metric
- multiple-choice accuracy in English; log-probability scoring
- Unit
- %
- Benchmark version
- Macaron (ACL 2026 Long Papers), Table 8; zero-shot evaluation
- Category
- multilingual and multicultural reasoning
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Macaron authors, ACL 2026 Long Papers, Table 8
- Status
- active
- Retrieved at
- 2026-10-19
- Methodology
- Independent evaluation. Aya Expanse-8B, zero-shot on Macaron; open-weight model. Scoring mode is log-probability scoring; Table 8 reports the exact row under model ID coherelabs-aya-expanse-8b.
- Context
- Table 8 reports accuracy in percent. The per-cell sample count, confidence interval, immutable checkpoint revision, and additional decoding parameters are not stated there; these values are the source-reported logprob/gen rows, not recomputed from rounded Table 3.
Page 1 of 5
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
Aya Expanse 8BGPT OSS 20B 41HUMAIN M3 41DeepSeek-V3.2 Exp 40Gemma 3n E2B IT 40
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 Aya Expanse 8B, 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
Frequently asked questions about Aya Expanse 8B
How large is the context window of Aya Expanse 8B?
8,000 tokens. That is how much text Aya Expanse 8B reads at once, about 10 A4 pages.
Page count as a rule of thumb from Artificial Analysis
GLM-5.1200K
DeepSeek-V3 0324164K
GLM-4.5131K
GPT OSS 120B131K
DeepSeek-V3.2128K
DeepSeek-R1128K
Aya Expanse 8BRank 103 of 1038K
Can I run Aya Expanse 8B locally?
Yes, the weights are open.
Stated by Cohere
- Open weights
- yes
- License
- CC BY-NC-4.0
- Parameters
- 8B
- active per token
- n/a
Sources and data date
Show all 21 sources
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
| Evaluator | Test series | Composite indices | Recorded active measurements |
|---|---|---|---|
| LMArena | 0 | 0 | 1 |
Without vendor claims and historical or disputed results; those stay labelled in the single values. 1 measurements are not yet assigned to a test series.