Compare · ModelsLive · 2 picked · head to head
Gemma 2 2b vs Vicuna 7b V1.5
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Vicuna 7b V1.5 wins on 3/6 benchmarks
Vicuna 7b V1.5 wins 3 of 6 shared benchmarks. Leads in general · language · reasoning.
Category leads
general·Vicuna 7b V1.5knowledge·Gemma 2 2blanguage·Vicuna 7b V1.5math·Gemma 2 2breasoning·Vicuna 7b V1.5
Hype vs Reality
Attention vs performance
Gemma 2 2b
#228 by perf·no signal
Vicuna 7b V1.5
#226 by perf·no signal
Vendor risk
Who is behind the model
Google DeepMind
$4.00T·Tier 1
L
LMSYS
private · undisclosed
Head to head
6 benchmarks · 2 models
Gemma 2 2bVicuna 7b V1.5
BBH (HuggingFace)
Vicuna 7b V1.5 leads by +2.7
Gemma 2 2b
12.5
Vicuna 7b V1.5
15.2
GPQA
Gemma 2 2b leads by +0.6
Gemma 2 2b
1.7
Vicuna 7b V1.5
1.1
IFEval
Vicuna 7b V1.5 leads by +3.3
Gemma 2 2b
20.2
Vicuna 7b V1.5
23.5
MATH Level 5
Gemma 2 2b leads by +1.7
Gemma 2 2b
3.0
Vicuna 7b V1.5
1.4
MMLU-PRO
Gemma 2 2b leads by +0.8
Gemma 2 2b
13.5
Vicuna 7b V1.5
12.7
MUSR
Vicuna 7b V1.5 leads by +0.2
Gemma 2 2b
11.3
Vicuna 7b V1.5
11.4
Full benchmark table
| Benchmark | Gemma 2 2b | Vicuna 7b V1.5 |
|---|---|---|
BBH (HuggingFace) | 12.5 | 15.2 |
GPQA | 1.7 | 1.1 |
IFEval | 20.2 | 23.5 |
MATH Level 5 | 3.0 | 1.4 |
MMLU-PRO | 13.5 | 12.7 |
MUSR | 11.3 | 11.4 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| Gemma 2 2b | — | — | — | — |
L Vicuna 7b V1.5 | — | — | — | — |