Note: Overall leaderboard rankings may not reflect true model quality — individual benchmarks give a clearer picture. ARC-Challenge MMLU GPQA GSM8K Artificial Analysis Intelligence Index v4.1
← Back to leaderboard

Amazon: MistralLite

4 benchmarks
CompanyAmazon
Country of originUSA

Leaderboard position

#771
Overall rank (of 1121 models)
531
Elo rating
4
Benchmarks
Top 68.8%
Percentile
#5
In Amazon (of 8 models)
Above: Qwen: qwen3-vl-4b-thinking #770 Below: Sapbot: resnet20-cifar10-272k #772

Benchmark highlights

Best benchmark
HellaSwag 81.617
Worst benchmark
TruthfulQA 38.275

Nearest competitor from another company

The model from another company with the most similar benchmark scores.

appeared as nearest in 1 of 4 benchmarks

Qwen: qwen-14b
1/4 25.0%
Qwen: qwen3-1.7b-base
1/4 25.0%
Apple: openelm-3b-instruct
1/4 25.0%
Qwen: qwen2.5-0.5b
1/4 25.0%
Model Votes % of benchmarks
Qwen: qwen-14b 1 25.0%
Qwen: qwen3-1.7b-base 1 25.0%
Apple: openelm-3b-instruct 1 25.0%
Qwen: qwen2.5-0.5b 1 25.0%

Benchmarks

HellaSwag 81.617 ARC-Challenge 58.874 MMLU 50.12 TruthfulQA 38.275

Predicted scores

Estimated from 25 similar models and the global Elo rating. These are projections, not measured results.

Benchmark Predicted Confidence Based on
ARC-Easy ~74.84 61% Google: gemma-3-4b-pt, Qwen: Qwen3.5 9B, IBM Granite: granite-4.1-3b
WinoGrande ~69.12 61% Google: gemma-3-4b-pt, Qwen: Qwen3.5 9B, IBM Granite: granite-4.1-3b
GPQA Diamond ~50.58 60% Qwen: Qwen3.5 9B, Microsoft: phi-4-mini-instruct, Hugging Face Smol Models Research: smollm3-3b
MMLU-Pro ~52.93 60% Google: gemma-3-4b-pt, Qwen: Qwen3.5 9B, IBM Granite: granite-4.1-3b
IFEval ~77.88 60% Qwen: Qwen3.5 9B, Meta Llama: Llama 3.2 3B Instruct, Hugging Face Smol Models Research: smollm3-3b
IFBench ~47.17 60% Microsoft: phi-4-mini-instruct, Meta Llama: Llama 3.2 3B Instruct, Hugging Face Smol Models Research: smollm3-3b
HumanEval ~63.21 60% Google: gemma-3-4b-pt, Technology Innovation Institute: falcon-h1-1.5b-deep-base, Qwen: qwen-7b
MBPP ~64.74 60% Google: gemma-3-4b-pt, Technology Innovation Institute: falcon-h1-1.5b-deep-base, Qwen: qwen-7b
GSM8K ~70.03 60% Google: gemma-3-4b-pt, IBM Granite: granite-4.1-3b, Technology Innovation Institute: falcon-h1-1.5b-deep-base
-(CoderPPL) ~-2.89 60% Qwen: Qwen3.5 9B, IBM Granite: granite-4.1-3b, Qwen: qwen3.5-9b-base
MATH ~47.19 60% Google: gemma-3-4b-pt, Qwen: qwen-7b, Qwen: qwen3-1.7b-base
BBH ~57.74 60% Technology Innovation Institute: falcon-h1-1.5b-deep-base, Qwen: qwen-7b, Technology Innovation Institute: falcon-h1-1.5b-base
PIQA ~80.38 60% Google: gemma-3-4b-pt, Apple: openelm-3b-instruct, Qwen: qwen-7b
CowQA ~57.76 60% Qwen: Qwen3.5 9B, DeepSeek: DeepSeek R1 Distill Llama 8B, Meta Llama: Llama 3.2 3B Instruct
BrowseComp ~31.13 60% NON AI - НЕ ИИ: Random baseline - Случайный счёт
GPQA ~38.99 60% Google: gemma-3-4b-pt, Technology Innovation Institute: falcon-h1-1.5b-deep-base, Technology Innovation Institute: falcon-h1-1.5b-base
Terminal-Bench v2.0 ~8.91 60% NON AI - НЕ ИИ: Random baseline - Случайный счёт
Artificial Analysis Intelligence Index v4.1 (Maximum Reasoning) ~11.1 60% Qwen: Qwen3.5 9B, Microsoft: phi-4-mini-instruct, DeepSeek: DeepSeek R1 Distill Llama 8B
Artificial Analysis Intelligence Index v4.0 (Maximum Reasoning) ~17.68 60% Qwen: Qwen3.5 9B, IBM Granite: granite-4.1-3b, Microsoft: phi-4-mini-instruct
Artificial Analysis Agentic Index (Maximum Reasoning) ~7.36 60% Qwen: Qwen3 0.6B
WikiText-2 (-ppl) ~-24.45 60% Qwen: Qwen3.5 9B, IBM Granite: granite-4.1-3b, Qwen: qwen3.5-9b-base
MATH-500 ~80.44 59% Microsoft: phi-4-mini-instruct, Hugging Face Smol Models Research: smollm3-3b, Technology Innovation Institute: falcon-h1-1.5b-instruct
AIME 2025 ~53.78 59% Microsoft: phi-4-mini-instruct, Hugging Face Smol Models Research: smollm3-3b, Technology Innovation Institute: falcon-h1-1.5b-instruct
Artificial Analysis Intelligence Index v4.3.2 ~8.74 59% Qwen: Qwen3.5 9B, Qwen: Qwen3.5 2B, LiquidAI: LFM2.5 1.2B Instruct
SuperGPQA ~32.37 59% Qwen: Qwen3.5 9B, Qwen: qwen3-1.7b-base, Google: gemma-3-1b-pt