NEWHow AI models draw the world map · Gecko Tests
Beta
Home/Models/Llama 3.2 90B
Meta logo

Llama 3.2 90B

by Meta · Released Jan 2024

Open Source
36.3
avg score
Rank #217
Compare
Better than 30% of all models
Context
N/A
Input $/1M
n/a
Output $/1M
n/a
Type
text
License
Open Source
Benchmarks
6 tested
Data as of
About

Tested on 6 benchmarks · BenchGecko score 36.3. Top scores: MMLU (73.7%), GeoBench (52.0%), MATH level 5 (39.4%).

Capabilities
math
21.0
#222 globally
knowledge
43.6
#175 globally
Benchmark Scores
Compare All
Tested on 6 benchmarks · Ranked across 2 categories
Score Distribution (all 312 models)
0255075100
▲ You are here
mathCompare math →
MATH level 5

Competition-level math from AMC, AIME, and olympiad problems. Level 5 is the hardest tier, requiring creative problem-solving.

39.4·
OTIS Mock AIME 2024-2025

Mock AIME (American Invitational Mathematics Exam) problems from OTIS. Tests mathematical competition performance.

2.5·
knowledgeCompare knowledge →
MMLU

Massive Multitask Language Understanding. 57 subjects from STEM, humanities, and social sciences. The most widely-cited knowledge benchmark.

73.7·
GeoBench

Geography benchmark testing knowledge of world geography, landmarks, borders, and geopolitical facts.

52.0·
Balrog

Broad Assessment of Language and Reasoning Over Games. Tests strategic and logical reasoning through game scenarios.

27.3·
Quick compare:Alibaba Qwen logovs Qwen3 8BGoogle DeepMind logovs Gemini 1.5 Pro (May 2024)Alibaba Qwen logovs Qwen2.5-Max
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Model Family · Meta Llama 3.2
Sep 2024
$0.34/M in131Kctx
Sep 2024
11.7
$0.03/M in(-0.32)60Kctx(-71K)11 benchmarks
Sep 2024
35.4+23.7
$0.05/M in(+0.02)131Kctx(+71K)7 benchmarks
Sep 2024
14.2-21.2
$0.00/M in(-0.05)131Kctx6 benchmarks
Jan 2024
36.3+22.1
n/an/actx6 benchmarks
See the full Llama 3.2 family →
Similar Models
Alibaba Qwen logo
Qwen3 8B
Alibaba Qwen
36.3$0.12/1M
Google DeepMind logo
Gemini 1.5 Pro (May 2024)
Google DeepMind
36.3n/a
Alibaba Qwen logo
Qwen2.5-Max
Alibaba Qwen
36.2n/a
Links
Info
MetaPricing explorerDevelopers · API
Research
Technical Report
Documentation
API DocsPlayground
Community
@Meta
Source Code
GitHubHugging Face
BenchGecko API
llama-3-2-90b
Specifications
  • Typetext
  • ContextN/A
  • ReleasedJan 2024
  • LicenseOpen Source
  • Statusbenchmark-only
Available On
Meta logoMetan/a
Categories
mathknowledge
Learn More
transformeropen weightstokens
Share & Export
Tweet
Related Models
Alibaba Qwen logoQwen3 8B
36.3
Google DeepMind logoGemini 1.5 Pro (May 2024)
36.3
Alibaba Qwen logoQwen2.5-Max
36.2
Meta logoLlama 2-13B
36.4
Llama 3.2 90B is an open-source text AI model by Meta, released in January 2024. It has an average benchmark score of 36.3.

Related Models

Qwen3 8B · Alibaba QwenGemini 1.5 Pro (May 2024) · Google DeepMindQwen2.5-Max · Alibaba QwenLlama 2-13B · MetaQwen3.5-Flash · Alibaba Qwen

Benchmarks

MMLUGeoBenchMATH level 5BalrogGPQA diamond

Related Pages

Meta · ProviderMeta · EconomyAll ModelsCompare ModelsPricingDevelopers · API

Key facts · as of 2026-04-09

  • Llama 3.2 90B by Meta. BenchGecko score 36.3, rank 216 of 312 scored models (normalized average of public benchmark scores).
  • List price n/a input · n/a output per 1M tokens (as of 2026-04-09).

How to cite · data as of 2026-04-09

Llama 3.2 90B · benchmarks, pricing and providers. BenchGecko, data as of 2026-04-09. https://benchgecko.ai/model/llama-3-2-90b

Credit "Source: BenchGecko" with a link. Prices per provider and Gecko Tests are BenchGecko data (CC BY 4.0); benchmark scores keep their original source, listed in the JSON. JSON · llms.txt · MCP