Model comparison
C4ai Aya Expanse 32b vs Llama 2-7B
C4ai Aya Expanse 32b is the stronger model overall, scoring 35.9 to 29.1 on the Noometry Index.
Last verified . 15 shared benchmarks.
Cohere
MetaSummary
- They share 15 benchmarks with published results for both. C4ai Aya Expanse 32b scores higher in 8 categories and Llama 2-7B in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in multilingual, where C4ai Aya Expanse 32b leads 38.4 to 23.8.
Side by side
| C4ai Aya Expanse 32b | Llama 2-7B | |
|---|---|---|
| Provider | Cohere | Meta |
| Noometry Index | 35.9 | 29.1 |
| Released | 2024-10-24 | 2023-07-18 |
| Weights | Open | Open |
| Context window | 128K | — |
| Max output | 4K | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 18 | 29 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding C4ai Aya Expanse 32b leads
C4ai Aya Expanse 32b: 34.8 (#231), Llama 2-7B: 29.2 (#307)
| Benchmark | C4ai Aya Expanse 32b | Llama 2-7B |
|---|---|---|
| LMArena Coding | 1197 | 1002 |
Reasoning C4ai Aya Expanse 32b leads
C4ai Aya Expanse 32b: 23.3 (#180), Llama 2-7B: 15.7 (#312)
| Benchmark | C4ai Aya Expanse 32b | Llama 2-7B |
|---|---|---|
| LMArena Hard Prompts | 1193 | 1009 |
| Chess Puzzles | — | 0% |
| BIG-Bench Hard | — | 39.2% |
| Epoch Capabilities Index | — | 99.06 |
| HellaSwag | — | 77.2% |
| LAMBADA | — | 73.3% |
| PIQA | — | 78.8% |
| WinoGrande | — | 69.2% |
Math C4ai Aya Expanse 32b leads
C4ai Aya Expanse 32b: 34.0 (#197), Llama 2-7B: 30.7 (#233)
| Benchmark | C4ai Aya Expanse 32b | Llama 2-7B |
|---|---|---|
| LMArena Math | 1200 | 1042 |
| GSM8K | — | 16.7% |
Knowledge C4ai Aya Expanse 32b leads
C4ai Aya Expanse 32b: 33.2 (#206), Llama 2-7B: 28.2 (#248)
| Benchmark | C4ai Aya Expanse 32b | Llama 2-7B |
|---|---|---|
| LMArena Expert | 1182 | 1036 |
| Vectara Hallucination Rate | 10.9% | — |
| ARC (AI2) Challenge | — | 45.9% |
| BoolQ | — | 77.9% |
| MMLU | — | 45.8% |
| OpenBookQA | — | 58.6% |
| TriviaQA | — | 73.7% |
Multimodal Not comparable
C4ai Aya Expanse 32b: —, Llama 2-7B: —
| Benchmark | C4ai Aya Expanse 32b | Llama 2-7B |
|---|---|---|
| ScienceQA | — | 43.1% |
Multilingual C4ai Aya Expanse 32b leads
C4ai Aya Expanse 32b: 38.4 (#230), Llama 2-7B: 23.8 (#293)
| Benchmark | C4ai Aya Expanse 32b | Llama 2-7B |
|---|---|---|
| LMArena Non-English | 1213 | 973 |
| LMArena Chinese | 1211 | 973 |
| LMArena French | 1248 | 970 |
| LMArena German | 1199 | 978 |
| LMArena Russian | 1227 | 995 |
| LMArena Spanish | 1193 | 1007 |
| LMArena Japanese | 1163 | — |
| LMArena Korean | 1158 | — |
Instruction Following C4ai Aya Expanse 32b leads
C4ai Aya Expanse 32b: 62.6 (#237), Llama 2-7B: 50.8 (#298)
| Benchmark | C4ai Aya Expanse 32b | Llama 2-7B |
|---|---|---|
| LMArena Instruction Following | 1196 | 1006 |
Long Context C4ai Aya Expanse 32b leads
C4ai Aya Expanse 32b: 37.2 (#220), Llama 2-7B: 30.4 (#287)
| Benchmark | C4ai Aya Expanse 32b | Llama 2-7B |
|---|---|---|
| LMArena Longer Query | 1228 | 999 |
Writing & Preference C4ai Aya Expanse 32b leads
C4ai Aya Expanse 32b: 42.2 (#235), Llama 2-7B: 28.0 (#298)
| Benchmark | C4ai Aya Expanse 32b | Llama 2-7B |
|---|---|---|
| LMArena Text | 1224 | 1053 |
| LMArena Creative Writing | 1200 | 1033 |
| LMArena Multi-Turn | 1190 | 1029 |
Frequently asked questions
Is C4ai Aya Expanse 32b better than Llama 2-7B?
C4ai Aya Expanse 32b is the stronger model overall, scoring 35.9 to 29.1 on the Noometry Index.
Is C4ai Aya Expanse 32b or Llama 2-7B better for coding?
C4ai Aya Expanse 32b scores higher on coding benchmarks: 34.8 versus 29.2 in the Noometry coding category.
How many benchmarks do C4ai Aya Expanse 32b and Llama 2-7B share?
15 benchmarks have published results for both models. C4ai Aya Expanse 32b has 18 scored results on Noometry and Llama 2-7B has 29.