whatmodel.sh / OPT-2.7B

OPT-2.7B

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

OPT-2.7B is an open-weights language model from Meta, released May 2022. Its published size is 2.7B parameters and a 2k-token context window. Licence: opt-175b-license-noncommercial.

Catalogue identity notes

Meta's Open Pre-trained Transformer base model as released on Hugging Face (weights under Meta's non-commercial OPT licence).

19.1±0.9#692 of 719 LLMsfrom 17 benchmark results across 3 sources

$ how it compares

The LLMs ranked either side of OPT-2.7B, out of 828 with a score, then more from Meta.

#modelscorewhereRAMreleased
796Flan-T5 Base Google20±3local~2.5GB2022-10
797Falcon-RW 1B TII20±2local~4.3GB2023-04
798OPT-2.7B Meta19±1local~7.1GB2022-05
799SmolLM2 360M Instruct Hugging Face19±3local~1.7GB2024-10
800GPT-Neo 2.7B EleutherAI18±1local~13GB2021-03
more from Meta
24Muse Spark 1.3 Meta77±1api $2.00/M2026-09
40Muse Spark 1.2 Meta75±1api $2.00/M2026-08
43Muse Spark 1.1 Meta75±1api $2.00/M2026-07

see all LLMs · find the ones that fit your machine

$ what it's good at

other tests11 not in a task · 100% of its score
testresultvs the others in its rundifficultycountswhy
ARC Challenge +131.3% ✎▁·▁··▁▃▇▂█▆▃ beats 3/26easy · avg 85%▇▇▇▇▇▇▇▇▇▇ 24%sharp · several runs
ARC Easy +154.3% ✎▁▁·▁▃▄▅▄█▅▁▃ beats 3/26very easy · avg 90%▇▇▇▇▇▇ 15%at its level · several runs
SciQ +185.8% ✎▁·▁·▃·▃▅▃█▄▁ beats 2/26very easy · avg 98%▇▇▇▇▇▇ 14%several runs
WinoGrande +161.0% ✎▁····▁··▁▃█▄ beats 5/26easy · avg 89%▇▇▇▇▇▇ 14%several runs
PIQA +173.8% ✎▁·▁··▁▁▃█▅█▂ beats 4/26easy · avg 88%▇▇▇▇ 8.8%blunt · at its level · several runs
BoolQ60.3% ✎▂▃▂▅▆▂▅▆█▂▂▂ beats 2/26easy · avg 91%▇▇▇ 7.2%at its level
LAMBADA (OpenAI version), last-word accuracy +163.6% ✎▁··▂▁▁▃▂▆█▅▃ beats 5/26very easy · avg 89%▇▇▇ 6.9%blunt · at its level · several runs
HellaSwag60.6% ✎▃···▁▇▅█▇▃▁▁ beats 4/26very easy · avg 92%▇▇ 5.8%at its level
3 more tests
testresultvs the others in its rundifficultycountswhy
Social IQA40.8% ✎▁···▁▃█▁▁▁·▁ beats 4/26medium · avg 59%▇ 1.6%blunt
OpenBookQA25.0% ✎█·▁▁▃·▁▃▅▇▃▄ beats 1/26medium · avg 53%▇ 1.4%too hard for it
Open PT-LLM Leaderboard average (9 Portuguese tasks)27.0% ✎▃▁▁▁▁▁▁▁▂█▅▁ beats 14/276medium · avg 73%▇ 0.6%noisy

$ where to run it

Other ways to run it

  1. Run it as published · needs ~5.3 GB · from huggingface.co/facebook
    uv run --with torch --with 'transformers<5' python -c "from transformers import pipeline; print(pipeline('text-generation', model='facebook/opt-2.7b')('Once upon a time', max_new_tokens=40)[0]['generated_text'])"

Weights and documentation: huggingface.co/facebook ↗

$ how the score was worked out · Stability AI — original Stabl… 65%, Pythia: A Suite for Analyzing… 34%, Open Portuguese LLM Leaderboa… 1%

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
Stability AI — original StableLM zero-shot evaluation outputs10▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 65%Apache-2.0
Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling — authors' own runs6▇▇▇▇▇▇▇▇▇▇ 34%CC BY-SA 4.0
Open Portuguese LLM Leaderboard (archived)1▇ 1%Apache-2.0
prior · pull toward the middle▇ 0%
weighted average= 19.1

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json