Live49 models · 46 open source · avg 28.5
Meta logo

Meta

🇺🇸United StatesWebsite
Top Model
Meta Llama 3 8B Instruct
avg score
49
Total Models
tracked on BenchGecko
46
Open Source
94% of models
$0.02
Cheapest Model
per 1M input tokens
28.5
Avg Benchmark
across 22 scored models

Model Categories

LLM44Multimodal5

Pricing Range · $/1M input tokens

$0.02
$0.03
$0.05
$0.10
$0.10
$0.14
$0.18
$0.19
$0.34
$0.40
$0.48
$0.51
Low: $0.02Median: $0.18High: $0.51

Open Source Ratio

94%
46 open source3 proprietary
#▲ModelAvgaider editaider poly?ANLI?APEX-Agents?ARC AI2?ARC-AGI?ARC-AGI-2?aa agentic?aa coding ?aa qualityaa critptaa gdpvalaa gpqa diaa humanitaa ifbenchaa long coaa mmmu praa scicodeaa tau2 beaa terminaseal audioseal audioseal audioBalrog?BBH?hf bbhC-EvalCadEval?charxiv re?charxiv re?arena elo arena elo chess puzz?Cl Benchcl bench lCMMLUCSQA2Cybench?deepresear?DeepsweDtbenchEbr BenchEnigmaEvalExploitbenchfiction li?FortressFrontiercodefrontierma?frontierma?frontiermafrontiermaFrontierswefurniture GdpvalGeoBench?GPQAGPQA diamond?graphwalks?GSM8K?GSO-Bench?HellaSwag?HELM — GPQAhelm ifevahelm mmlu helm omni helm wildbHLE?hle toolsseal humanseal humanIFEvaljp jcommonJHumanEvalJMMLUJNLIJSQuADLAMBADA?lech mazur?livebench livebench livebench livebench livebench livebench livebench livebench jp overallLmcaMASKMATH level 5?MATH Level 5MCP Atlasmetr time MirrorcodeMMLU?MMLU-PROMMMLUmmmlu armmmlu bnmmmlu zhmmmlu frmmmlu demmmlu himmmlu idmmmlu itmmmlu jammmlu kommmlu ptmmmlu esmmmlu swmmmlu yoseal multiMultiNRCMUSRmystery gaOpenBookQA?oc aime202oc gpqa dioc hleoc ifevaloc livecodoc mmlu prOSWorld?Osworld 2 0otis mock ?PIQA?posttrainbseal pro rseal pro rProofbenchseal properemote labseal remotScienceQA?SciPredictSimpleBench?simpleqa v?surface evseal swe aseal swe aswe bench swe bench swe bench seal swe bseal swe bswe bench ?swe bench ?terminal b?the agent ?TriviaQA?TutorBenchUSAMOVideoMME?VISTAseal visuaVPCT?WeirdML?Winogrande?$/1M inContextReleased
1Meta logoMeta Llama 3 8B🇺🇸 MetaOpen64.0-------------------------48.7-----------------------------19.7--------------16.082.9-44.760.988.9----------48.9---18.6----41.2-----------------16.0--------------------------------------------n/a-Apr 242y ago
2Meta logoLlama 3.1 70B Instruct🇺🇸 MetaOpen49.758.6----------------------27.9-55.9-----1293.3---64.4----33.3--------------14.225.6-------------86.7----------------17.5-36.738.1---73.547.9-----------------17.7----------3.5-----------------------6.9-------9.0-$0.40131KJul 242y ago
3Meta logoLlama 3.3 70B Instruct🇺🇸 MetaOpen47.959.4----------------------23.0-56.6-----1317.8--------32.5---33.3----------10.529.9-------------90.0----------------20.6-41.648.3---81.748.1-----------------15.6----------5.0----------3.9--------------------14.4-$0.10131KDec 241y ago
4Meta logoLlama 3.1 405B🇺🇸 MetaOpen43.8----93.7-------------------77.27.8-----------7.5--35.6--------------5.934.5---85.6---------18.1------------------49.80.0---79.325.7-----------------2.2-32.3--------9.671.8---------7.6------------7.482.7------21.478.4n/a-Jul 242y ago
5Meta logoLlama 3 8B Instruct🇺🇸 MetaOpen41.8--36.0-77.1--------------------18.4-----1222.8-----------------------2.11.4-------------24.0------------------6.13.9---58.417.8-----------------19.9-76.8--------0.7------------------------67.7-------51.4$0.148KApr 242y ago
6Meta logoLlama 3-70B🇺🇸 MetaOpen41.6--36.0-77.1--------------------26.7------0.0----5.0--23.6--------------8.120.8-------------33.2------------------22.65.7---72.424.0-----------------7.1-76.8--------4.2------------------------67.7-------67.0n/a-Jan 242y ago
7Meta logoLlama 2-13B🇺🇸 MetaOpen36.4----47.1-------------------44.3-------0.0---0.1---3.7---------------1.8-36.9-74.3---------------76.5------------3.3----40.8--------------------42.7--------0.061.6-------41.0---------------79.6-------45.6n/a-Jan 242y ago
8Meta logoLlama 3.2 90B🇺🇸 MetaOpen36.3-----------------------27.3------------------------------52.0-21.4--------------------------------39.4----73.7-----------------------------2.5---------------------------------n/a-Jan 242y ago
9Meta logoLlama 3.2 3B Instruct🇺🇸 MetaOpen35.4-------------------------24.1-----1166.7-----------------------3.8--------------73.9-------------------17.7----24.4-----------------1.4--------------------------------------------$0.05131KSep 242y ago
10Meta logoLLaMA-13B🇺🇸 MetaOpen33.4----36.9-------------------17.225.338.8----975.1---39.8-------------------3.5--20.6-72.3---------25.3-----75.2-------------3.1---30.323.1-----------------2.0-41.9---------60.2-------24.4---------------77.9-------46.0n/a-Jan 242y ago
11Meta logoMeta Llama 3 8B Instruct🇺🇸 MetaOpen30.7-------------------------28.2-----------------------------1.2--------------74.187.7-46.761.189.5----------49.6---8.7----29.6-40.536.451.455.853.541.451.053.342.346.555.555.837.531.0--1.6--------------------------------------------n/a-Apr 242y ago
12Meta logoLlama 3.1 8B Instruct🇺🇸 MetaOpen30.737.6----------------------15.1-29.2-----1211.20.0-------18.2--------------9.52.6-82.4-----------50.6----------------6.3-22.915.5---41.530.9-----------------8.5----------1.662.4------------------------------1.7-$0.02131KJul 242y ago
13Meta logoLlama 3 70B Instruct🇺🇸 MetaOpen30.6-------------------------------1275.1---36.9-5.0------------------20.8--------------------------------22.6----72.4--------------------30.1--------4.2--------------------------------67.0$0.518KApr 242y ago
14Meta logoopen_llama_7b🇺🇸 MetaOpen30.0----18.3-------------------------------------------------------62.4---------------------------------6.5--------------------18.7---------52.0-------------------------------34.0n/a-Jan 242y ago
15Meta logoLlama 3.3 70B Instruct (free)🇺🇸 MetaOpen28.9-----------------------23.0--------------------33.3-----------29.9--------------------------------41.6----81.7-----------------------------5.0----------3.9--------------------14.4-Free131KDec 241y ago
16Meta logoLlama 2 7b Chat Hf🇺🇸 MetaOpen25.4-------------------------4.5-----------------------------0.5--------------39.952.6-33.335.683.0----------38.3---2.0----7.6-----------------3.3--------------------------------------------n/a-Jul 233y ago
17Meta logoLlama 4 Maverick🇺🇸 MetaOpen22.1-15.6---4.40.11.316.310.00.00.067.14.943.050.062.131.717.86.8--------------------36.5---46.2--1.2------52.0-56.0---------0.9----------62.0---------18.7-73.0----------------------------------20.5----------13.2----------21.0---------24.5-$0.191.0MApr 251y ago
18Meta logoLlama 2 7b Hf🇺🇸 MetaOpen21.5-------------------------10.3-----------------------------2.2--------------25.225.5-28.636.179.9----------37.2---1.7----9.6-----------------3.8--------------------------------------------n/a-Jul 233y ago
19Meta logoLlama 4 Scout🇺🇸 MetaOpen16.5-----0.50.11.18.28.10.00.058.73.839.527.752.921.315.51.5--------------------29.8---36.0--0.0--------35.8------------------------------14.1-62.3----------------------------------7.7---------------------9.1-----------$0.101.3MApr 251y ago
20Meta logoLlama 3.2 3B Instruct (free)🇺🇸 MetaOpen14.2-------------------------14.2-----------------------------2.4--------------13.4-------------------1.9----16.5-----------------3.8--------------------------------------------Free131KSep 242y ago
21Meta logoLlama 3.2 1B Instruct🇺🇸 MetaOpen11.7-----------------------6.6-8.3-----1111.10.0----------------------2.40.0-------------58.1-------------------8.2----8.2-----------------1.9----------0.5---------------------------------$0.0360KSep 242y ago
90+ Gold 80-89 70-79 60-69 <60Scores in % unless noted. Avg = unweighted mean across tested benchmarks.

Quick answers · sourced from our data

How many models does Meta have?

BenchGecko tracks 49 models from Meta, of which 46 (94%) are open source. Every entry is updated daily from live provider feeds.

What is the best model from Meta?

Meta Llama 3 8B Instruct is currently the highest scoring Meta model we track, with an average benchmark score of 45.2. Scores are computed across every public benchmark we have data for.

What is the cheapest Meta model?

The cheapest Meta model on BenchGecko starts at $0.02 per 1M input tokens. Pricing is pulled from OpenRouter and cross-checked against official provider rate cards.

How does Meta compare on benchmarks?

Meta models average 28.5 across the benchmarks we track · see the All Providers page for the full ranking by model count, open source ratio, and average score.

Where is Meta based?

Meta is headquartered in United States. BenchGecko groups providers by region to make it easy to compare US, EU, China, and Rest of World markets.

Is Meta open source?

46 of 49 Meta models are open source (94%). The rest are proprietary · closed weights served via API.