Comparisons of Tiny Open Source AI Models (≤4B)
Open source AI models with 4B parameters or fewer. These are usually the smallest models in terms of resource demand. Models are considered Open Source (also commonly referred to as open weights) where their weights are accessible to download. This allows self-hosting on your own infrastructure and enables customizing the model such as through fine-tuning. Click on any model to see detailed metrics. For more details including relating to our methodology, see our FAQs.
Qwen3 4B 2507 and Qwen3 VL 4B are the highest intelligence Tiny open source models, defined as those with ≤4B parameters, followed by Qwen3 4B 2507 & 
Ministral 3 3B.
Highlights
Openness
Artificial Analysis Openness Index: Results
Intelligence
Artificial Analysis Intelligence Index
Artificial Analysis Intelligence Index v4.0 includes: GDPval-AA, 𝜏²-Bench Telecom, Terminal-Bench Hard, SciCode, AA-LCR, AA-Omniscience, IFBench, Humanity's Last Exam, GPQA Diamond, CritPt. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them.
{"@context":"https://schema.org","@type":"Dataset","name":"Artificial Analysis Intelligence Index","creator":{"@type":"Organization","name":"Artificial Analysis","url":"https://artificialanalysis.ai"},"description":"Artificial Analysis Intelligence Index v4.0 incorporates 10 evaluations: GDPval-AA, 𝜏²-Bench Telecom, Terminal-Bench Hard, SciCode, AA-LCR, AA-Omniscience, IFBench, Humanity's Last Exam, GPQA Diamond, CritPt","measurementTechnique":"Independent test run by Artificial Analysis on dedicated hardware.","spatialCoverage":"Worldwide","keywords":["analytics","llm","AI","benchmark","model","gpt","claude"],"license":"https://creativecommons.org/licenses/by/4.0/","isAccessibleForFree":true,"citation":"Artificial Analysis (2025). LLM benchmarks dataset. https://artificialanalysis.ai","data":""}
Intelligence Evaluations
While model intelligence generally translates across use cases, specific evaluations may be more relevant for certain use cases.
Artificial Analysis Intelligence Index v4.0 includes: GDPval-AA, 𝜏²-Bench Telecom, Terminal-Bench Hard, SciCode, AA-LCR, AA-Omniscience, IFBench, Humanity's Last Exam, GPQA Diamond, CritPt. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them.
Model Size: Total and Active Parameters
The total number of trainable weights and biases in the model, expressed in billions. These parameters are learned during training and determine the model's ability to process and generate responses.
The number of parameters actually executed during each inference forward pass, expressed in billions. For Mixture of Experts (MoE) models, a routing mechanism selects a subset of experts per token, resulting in fewer active than total parameters. Dense models use all parameters, so active equals total.
Intelligence vs. Active Parameters
Artificial Analysis Intelligence Index v4.0 includes: GDPval-AA, 𝜏²-Bench Telecom, Terminal-Bench Hard, SciCode, AA-LCR, AA-Omniscience, IFBench, Humanity's Last Exam, GPQA Diamond, CritPt. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them.
The number of parameters actually executed during each inference forward pass, expressed in billions. For Mixture of Experts (MoE) models, a routing mechanism selects a subset of experts per token, resulting in fewer active than total parameters. Dense models use all parameters, so active equals total.
Intelligence vs. Total Parameters
Artificial Analysis Intelligence Index v4.0 includes: GDPval-AA, 𝜏²-Bench Telecom, Terminal-Bench Hard, SciCode, AA-LCR, AA-Omniscience, IFBench, Humanity's Last Exam, GPQA Diamond, CritPt. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them.
The total number of trainable weights and biases in the model, expressed in billions. These parameters are learned during training and determine the model's ability to process and generate responses.
Context Window
Context Window
Larger context windows are relevant to RAG (Retrieval Augmented Generation) LLM workflows which typically involve reasoning and information retrieval of large amounts of data.
Maximum number of combined input & output tokens. Output tokens commonly have a significantly lower limit (varied by model).
{"@context":"https://schema.org","@type":"Dataset","name":"Context Window","creator":{"@type":"Organization","name":"Artificial Analysis","url":"https://artificialanalysis.ai"},"description":"Context Window: Tokens Limit; Higher is better","measurementTechnique":"Independent test run by Artificial Analysis on dedicated hardware.","spatialCoverage":"Worldwide","keywords":["analytics","llm","AI","benchmark","model","gpt","claude"],"license":"https://creativecommons.org/licenses/by/4.0/","isAccessibleForFree":true,"citation":"Artificial Analysis (2025). LLM benchmarks dataset. https://artificialanalysis.ai","data":""}
| Weights | Provider Benchmarks | |||||||
|---|---|---|---|---|---|---|---|---|
Qwen3 4B 2507 (Reasoning) Alibaba | 19 | 4.02B | 262k | - | - | 🤗 | - | View |
Qwen3 VL 4B (Reasoning) Alibaba | 15 | 4.44B | 256k | - | - | 🤗 | - | View |
Qwen3 4B 2507 Instruct Alibaba | 13 | 4.02B | 262k | - | - | 🤗 | - | View |
 Ministral 3 3B Mistral | 11 | 3B | 256k | $0.1 | 314 | 🤗 |  | View |
 Phi-4 Mini Instruct Microsoft Azure | 11 | 3.84B | 128k | - | 45 | 🤗 | | View |
 Jamba Reasoning 3B AI21 Labs | 10 | 3B | 262k | - | - | 🤗 | - | View |
Qwen3 VL 4B Instruct Alibaba | 10 | 4.44B | 256k | - | - | 🤗 | - | View |
 Gemma 3 1B Instruct Google | 9 | 1B | 32.0k | - | 44 | 🤗 | | View |
Gemma 3 270M Google | 8 | 0.268B | 32.0k | - | - | 🤗 | - | View |
 Exaone 4.0 1.2B (Reasoning) LG AI Research | 8 | 1.28B | 64.0k | - | - | 🤗 | - | View |
 LFM2.5-1.2B-Thinking Liquid AI | 8 | 1.17B | 32.0k | - | - | 🤗 | - | View |
Exaone 4.0 1.2B (Non-reasoning) LG AI Research | 8 | 1.28B | 64.0k | - | - | 🤗 | - | View |
 Granite 4.0 H 1B IBM | 8 | 1.5B | 128k | - | - | 🤗 | - | View |
LFM2.5-1.2B-Instruct Liquid AI | 8 | 1.17B | 32.0k | - | - | 🤗 | ? | View |
Qwen3 1.7B (Reasoning) Alibaba | 8 | 2.03B | 32.0k | $0.4 | 132 | 🤗 | | View |
LFM2 2.6B Liquid AI | 8 | 2.57B | 32.8k | - | - | 🤗 | ? | View |
Granite 4.0 Micro IBM | 8 | 3B | 128k | - | - | 🤗 | - | View |
Granite 4.0 1B IBM | 7 | 1.6B | 128k | - | - | 🤗 | - | View |
Qwen3 1.7B (Non-reasoning) Alibaba | 7 | 2.03B | 32.0k | $0.2 | 122 | 🤗 | | View |
Granite 4.0 350M IBM | 7 | 0.35B | 32.8k | - | - | 🤗 | - | View |
Qwen3 0.6B (Reasoning) Alibaba | 6 | 0.752B | 32.0k | $0.4 | 211 | 🤗 | | View |
Gemma 3 4B Instruct Google | 6 | 4.3B | 128k | - | 43 | 🤗 |  | View |
LFM2.5-VL-1.6B Liquid AI | 6 | 1.6B | 32.0k | - | - | 🤗 | - | View |
Qwen3 0.6B (Non-reasoning) Alibaba | 6 | 0.752B | 32.0k | $0.2 | 192 | 🤗 | | View |
Granite 4.0 H 350M IBM | 5 | 0.34B | 32.8k | - | - | 🤗 | - | View |
 DeepSeek-OCR DeepSeek | - | 3.34B | 8.19k | $0.0 | 341 | 🤗 |  | View |