MODEL DIRECTORY
Follow the model.
See the evidence.
Benchmark results, evaluation settings and disclosed training methods—collected in one place for each model.
3,335 reported models and checkpoint variants. Scores keep their original units and sources.
662 models · page 4 of 14
Meta
OPT-66B
3 benchmarks · 3 observations
Explore results →Meta
OPT-175B
3 benchmarks · 3 observations
Explore results →PaLM (540B)
12 benchmarks · 31 observations
Explore results →GLaM
4 benchmarks · 12 observations
Explore results →NVIDIA
Megatron-Turing NLG 530B
5 benchmarks · 19 observations
Explore results →T5 (11B)
4 benchmarks · 5 observations
Explore results →T5-3B
3 benchmarks · 3 observations
Explore results →Gemini 1.5 Pro
32 benchmarks · 51 observations
Explore results →Gemini 3.1 Pro Preview (02/26)
32 benchmarks · 35 observations
Explore results →Gemini 3.1 Flash Lite Preview
27 benchmarks · 28 observations
Explore results →Gemini 3 Flash (12/25)
25 benchmarks · 26 observations
Explore results →Gemini 1.5 Flash
22 benchmarks · 32 observations
Explore results →Gemini 3 Pro (11/25)
18 benchmarks · 19 observations
Explore results →Gemini 1.5 Pro (002)
17 benchmarks · 37 observations
Explore results →Meta
Llama 3.1-8B-Instruct
17 benchmarks · 32 observations
Explore results →Gemini 1.5 Flash (002)
16 benchmarks · 17 observations
Explore results →Meta
Llama 3-70B-Instruct
15 benchmarks · 50 observations
Explore results →Gemini 2.5 Flash Lite (9/25)
13 benchmarks · 28 observations
Explore results →Gemini 2.5 Flash Preview (9/25)
13 benchmarks · 28 observations
Explore results →Meta
Llama 3-8B-Instruct
13 benchmarks · 47 observations
Explore results →NVIDIA
Nemotron 3.5 Lightning
13 benchmarks · 13 observations
Explore results →Gemini 2.0 Flash (001)
12 benchmarks · 13 observations
Explore results →Meta
Llama 3.1-70B-Instruct
12 benchmarks · 26 observations
Explore results →gemini-2-0-flash
10 benchmarks · 11 observations
Explore results →gemini-2.5-flash*
10 benchmarks · 16 observations
Explore results →Meta
Llama 2-70b / chat
10 benchmarks · 24 observations
Explore results →Gemini 2.5 Pro Exp
9 benchmarks · 10 observations
Explore results →Gemini 2.5 Flash Preview 4/17
8 benchmarks · 15 observations
Explore results →Meta
Llama 3 405B
8 benchmarks · 8 observations
Explore results →Meta
Llama 3.3 Instruct Turbo (70B)
8 benchmarks · 8 observations
Explore results →Meta
Llama 2-7b / chat
7 benchmarks · 12 observations
Explore results →Meta
Llama 3-8B Base
7 benchmarks · 7 observations
Explore results →Meta
Llama 3.1 405B Instruct
7 benchmarks · 33 observations
Explore results →Meta
Llama 3.3 70B Instruct
7 benchmarks · 20 observations
Explore results →Meta
CodeLlama-13B‡
6 benchmarks · 6 observations
Explore results →Meta
CodeLlama-34B‡
6 benchmarks · 6 observations
Explore results →Meta
CodeLlama-7B‡
6 benchmarks · 6 observations
Explore results →gemini-2.5-pro-preview-06-05
6 benchmarks · 6 observations
Explore results →Gemma-7b-it
6 benchmarks · 10 observations
Explore results →Meta
Llama 3.1-70B-Inst
6 benchmarks · 11 observations
Explore results →NVIDIA
Llama 3.3 Nemotron Super
6 benchmarks · 12 observations
Explore results →Meta
Muse Wonder 1.0
6 benchmarks · 7 observations
Explore results →Gemini 2.5 Flash Lite
5 benchmarks · 6 observations
Explore results →Gemini Pro 1.5
5 benchmarks · 5 observations
Explore results →gemini-1.5-pro-001
5 benchmarks · 41 observations
Explore results →Gemini-pro
5 benchmarks · 10 observations
Explore results →gemma-1.1-7b-it
5 benchmarks · 7 observations
Explore results →Meta
Llama 2-13b / chat
5 benchmarks · 15 observations
Explore results →