Comparison of Open Source Models

Name: Openness
Creator: Artificial Analysis
License: https://artificialanalysis.ai/docs/legal/Terms-of-Use.pdf

Comparison and analysis of open source AI models across key performance metrics including quality, performance, inference speed, context window, parameter count & licensing details. 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 relating to our methodology, see our FAQs.

Kimi K2.6 and

MiMo-V2.5-Pro are the highest intelligence open source models, followed by

DeepSeek V4 Pro (Max) &

GLM-5.1.

Highlights

Openness

Artificial Analysis Openness Index · Higher is better

Intelligence

Artificial Analysis Intelligence Index · Higher is better

Total Parameters

Trainable parameters in billions

Openness

Artificial Analysis Openness Index: Results

Openness Index assesses model openness on a 0 to 100 normalized scale (higher is more open)

Reasoning models are indicated by a lightbulb icon

Open Source Progress

Progress in Open Weights vs. Proprietary Intelligence

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

Reasoning models are indicated by a lightbulb icon.

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.

Indicates whether the model weights are available. Models are labelled as 'Commercial Use Restricted' if the weights are available but commercial use is limited (typically requires obtaining a paid license).

Open Source Language Models Intelligence By Lab Over Time

Reasoning models are indicated by a lightbulb icon.

Open Source Models Intelligence By Size Over Time

Reasoning models are indicated by a lightbulb icon.

Tiny: Less than or equal to 4B parameters. These are usually the smallest models in terms of resource demand.
Small: Less than 40B parameters.
Medium: Between 40B-150B parameters.
Large: Over 150B parameters.

Intelligence

Artificial Analysis Intelligence Index

Estimate (independent evaluation forthcoming)

Reasoning models are indicated by a lightbulb icon

Intelligence Evaluations

Intelligence evaluations measured independently by Artificial Analysis · Higher is better

GDPval-AA

Agentic real-world work tasks, (Elo-500)/2000

Terminal-Bench Hard

Agentic coding & terminal use

𝜏²-Bench Telecom

Agentic tool use

AA-LCR

Long context reasoning

AA-Omniscience Accuracy

Knowledge

AA-Omniscience Non-Hallucination Rate

1 - hallucination rate

Humanity's Last Exam

Reasoning & knowledge

GPQA Diamond

Scientific reasoning

SciCode

Coding

IFBench

Instruction following

CritPt

Physics reasoning

APEX-Agents-AA

Long-horizon agentic tasks

ITBench-AA

Kubernetes incident root-cause analysis

MMMU-Pro

Visual reasoning

Reasoning models are indicated by a lightbulb icon.

While model intelligence generally translates across use cases, specific evaluations may be more relevant for certain use cases.

Size

Intelligence Index By Model Size

Estimate (independent evaluation forthcoming)

Large Models (>150B)

Medium Models (40B-150B)

Small Models (4B-40B)

Reasoning models are indicated by a lightbulb icon.

Tiny: Less than or equal to 4B parameters. These are usually the smallest models in terms of resource demand.
Small: Less than 40B parameters.
Medium: Between 40B-150B parameters.
Large: Over 150B parameters.

Model Size: Total and Active Parameters

Comparison between total model parameters and parameters active during inference

Reasoning models are indicated by a lightbulb icon

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

Active parameters at inference time · Artificial Analysis Intelligence Index

Most attractive quadrant

Alibaba

DeepSeek

Google

Kimi

MBZUAI Institute of Foundation Models

MiniMax

Mistral

NVIDIA

OpenAI

Xiaomi

Z AI

Reasoning models are indicated by a lightbulb icon.

Intelligence vs. Total Parameters

Artificial Analysis Intelligence Index · Size in parameters (billions)

Most attractive quadrant

Alibaba

DeepSeek

Google

Kimi

MBZUAI Institute of Foundation Models

MiniMax

Mistral

NVIDIA

OpenAI

Xiaomi

Z AI

Reasoning models are indicated by a lightbulb icon.

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: tokens limit · Higher is better

Reasoning models are indicated by a lightbulb icon

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).

Further details

							Weights		Provider Benchmarks
Kimi K2.6	Kimi	54	1.0KB 32B active at inference time	256k	$0.7	57	🤗	+12	View
MiMo-V2.5-Pro	Xiaomi	54	1.0KB 42B active at inference time	1.00M	$0.2	41	🤗		View
DeepSeek V4 Pro (Reasoning, Max Effort)	DeepSeek	52	1.6KB 49B active at inference time	1.00M	$0.2	61	🤗	+8	View
GLM-5.1 (Reasoning)	Z AI	51	744B 40B active at inference time	200k	$0.9	83	🤗	+9	View
DeepSeek V4 Pro (Reasoning, High Effort)	DeepSeek	50	1.6KB 49B active at inference time	1.00M	$0.2	61	🤗	+8	View
GLM-5 (Reasoning)	Z AI	50	744B 40B active at inference time	200k	$0.7	73	🤗	+9	View
MiniMax-M2.7	MiniMax	50	230B 10B active at inference time	205k	$0.2	59	🤗	+3	View
MiMo-V2.5	Xiaomi	49	310B 15B active at inference time	1.00M	$0.1	76	🤗	+2	View
Nemotron 3 Ultra 550B A55B (Reasoning)	NVIDIA	48	550B 55B active at inference time	262k	$0.5	163	Not available	+4	View
Kimi K2.5 (Reasoning)	Kimi	47	1.0KB 32B active at inference time	256k	$0.6	44	🤗	+12	View
DeepSeek V4 Flash (Reasoning, Max Effort)	DeepSeek	47	284B 13B active at inference time	1.00M	$0.1	100	🤗	+4	View
DeepSeek V4 Flash (Reasoning, High Effort)	DeepSeek	46	284B 13B active at inference time	1.00M	$0.1	-	🤗	+4	View
Qwen3.6 27B (Reasoning)	Alibaba	46	27.8B	262k	$0.9	61	🤗	+2	View
Qwen3.5 397B A17B (Reasoning)	Alibaba	45	397B 17B active at inference time	262k	$0.9	52	🤗	+9	View
GLM-5.1 (Non-reasoning)	Z AI	44	744B 40B active at inference time	200k	$0.9	80	🤗	+5	View
Qwen3.6 35B A3B (Reasoning)	Alibaba	43	36B 3B active at inference time	262k	$0.4	179	🤗	+6	View
Kimi K2.6 (Non-reasoning)	Kimi	43	1.0KB 32B active at inference time	256k	$0.7	45	🤗	+9	View
Step 3.7 Flash	StepFun	43	198B 11B active at inference time	256k	$0.2	186	🤗		View
GLM-4.7 (Reasoning)	Z AI	42	357B 32B active at inference time	200k	$0.7	76	🤗	+7	View
Qwen3.5 27B (Reasoning)	Alibaba	42	27.8B	262k	$0.5	83	🤗	+3	View
MiniMax-M2.5	MiniMax	42	230B 10B active at inference time	205k	$0.3	215	🤗	+13	View
Hy3-preview (Reasoning)	Tencent	42	295B 21B active at inference time	256k	$0.1	93	🤗		View
DeepSeek V3.2 (Reasoning)	DeepSeek	42	685B 37B active at inference time	128k	$0.2	-	🤗	+12	View
Qwen3.5 122B A10B (Reasoning)	Alibaba	42	125B 10B active at inference time	262k	$0.7	137	🤗	+2	View
MiMo-V2-Flash (Feb 2026)	Xiaomi	41	309B 15B active at inference time	256k	$0.1	143	🤗		View
Kimi K2 Thinking	Kimi	41	1.0KB 32B active at inference time	256k	$0.8	124	🤗	+3	View
GLM-5 (Non-reasoning)	Z AI	41	744B 40B active at inference time	200k	$0.7	71	🤗	+3	View
Qwen3.5 397B A17B (Non-reasoning)	Alibaba	40	397B 17B active at inference time	262k	$0.9	53	🤗	+6	View
MiniMax-M2.1	MiniMax	39	230B 10B active at inference time	205k	$0.4	232	🤗		View
DeepSeek V4 Pro (Non-reasoning)	DeepSeek	39	1.6KB 49B active at inference time	1.00M	$0.2	60	🤗	+2	View
MiMo-V2-Flash (Reasoning)	Xiaomi	39	309B 15B active at inference time	256k	$0.1	150	🤗		View
Mistral Medium 3.5	Mistral	39	128B	256k	$2.1	147	🤗		View
Gemma 4 31B (Reasoning)	Google	39	30.7B	256k	-	35	🤗	+8	View
Ring-2.6-1T	InclusionAI	38	1.0KB 63B active at inference time	262k	$0.5	127	🤗		View
Step 3.5 Flash	StepFun	38	196B 11B active at inference time	256k	$0.1	175	🤗		View
Kimi K2.5 (Non-reasoning)	Kimi	37	1.0KB 32B active at inference time	256k	$0.8	46	🤗	+6	View
Qwen3.5 27B (Non-reasoning)	Alibaba	37	27.8B	262k	$0.5	91	🤗		View
Command A+	Cohere	37	218B 25B active at inference time	192k	-	200	🤗		View
Qwen3.6 27B (Non-reasoning)	Alibaba	37	27.8B	262k	$0.9	59	🤗		View
Qwen3.5 35B A3B (Reasoning)	Alibaba	37	36B 3B active at inference time	262k	$0.4	154	🤗	+2	View
DeepSeek V4 Flash (Non-reasoning)	DeepSeek	36	284B 13B active at inference time	1.00M	$0.1	97	🤗		View
MiniMax-M2	MiniMax	36	230B 10B active at inference time	205k	$0.4	113	🤗		View
NVIDIA Nemotron 3 Super 120B A12B (Reasoning)	NVIDIA	36	120.6B 12.7B active at inference time	1.00M	$0.3	154	🤗	+2	View
Qwen3.5 122B A10B (Non-reasoning)	Alibaba	36	125B 10B active at inference time	262k	$0.7	163	🤗		View
MiMo-V2.5-Pro (Non-reasoning)	Xiaomi	36	1.0KB 41.7B active at inference time	1.00M	$0.6	56	🤗		View
GLM-4.7 (Non-reasoning)	Z AI	34	357B 32B active at inference time	200k	$0.7	76	🤗	+6	View
DeepSeek V3.1 Terminus (Reasoning)	DeepSeek	34	685B 37B active at inference time	128k	$1.7	-	🤗		View
Hy3-preview (Non-reasoning)	Tencent	34	295B 21B active at inference time	256k	$0.1	93	🤗		View
Ling-2.6-1T	InclusionAI	34	1.0KB 63B active at inference time	262k	$0.5	-	🤗		View
gpt-oss-120b (high)	OpenAI	33	117B 5.1B active at inference time	131k	$0.2	353	🤗	+23	View
DeepSeek V3.2 Exp (Reasoning)	DeepSeek	33	685B 37B active at inference time	128k	$0.2	-	🤗		View
GLM-4.6 (Reasoning)	Z AI	33	357B 32B active at inference time	200k	$0.7	50	🤗		View
Qwen3.5 9B (Reasoning)	Alibaba	32	9.65B	262k	$0.1	66	🤗		View
Gemma 4 31B (Non-reasoning)	Google	32	30.7B	256k	$0.2	44	🤗	+4	View
K-EXAONE (Reasoning)	LG AI Research	32	236B 23B active at inference time	256k	-	-	🤗	-	View
DeepSeek V3.2 (Non-reasoning)	DeepSeek	32	685B 37B active at inference time	128k	$0.5	-	🤗	+12	View
Trinity Large Thinking	Arcee AI	32	399B 13B active at inference time	512k	$0.2	211	🤗		View
Qwen3.6 35B A3B (Non-reasoning)	Alibaba	32	36B 3B active at inference time	262k	$0.6	192	🤗	+5	View
Gemma 4 26B A4B (Reasoning)	Google	31	25.2B 3.8B active at inference time	256k	$0.1	-	🤗	+4	View
Kimi K2 0905	Kimi	31	1.0KB 32B active at inference time	256k	$0.8	25	🤗		View
Qwen3.5 35B A3B (Non-reasoning)	Alibaba	31	36B 3B active at inference time	262k	$0.4	184	🤗		View
MiMo-V2-Flash (Non-reasoning)	Xiaomi	30	309B 15B active at inference time	256k	$0.1	149	🤗		View
GLM-4.6 (Non-reasoning)	Z AI	30	357B 32B active at inference time	200k	$0.8	53	🤗		View
EXAONE 4.5 33B	LG AI Research	30	34.4B	262k	-	-	🤗	-	View
GLM-4.7-Flash (Reasoning)	Z AI	30	31.2B 3B active at inference time	200k	$0.1	73	🤗		View
Qwen3 235B A22B 2507 (Reasoning)	Alibaba	30	235B 22B active at inference time	256k	$0.6	60	🤗	+3	View
DeepSeek V3.2 Speciale	DeepSeek	29	685B 37B active at inference time	128k	-	-	🤗	-	View
Gemma 4 12B (Reasoning)	Google	29	12B	256k	-	-	🤗	-	View
DeepSeek V3.1 Terminus (Non-reasoning)	DeepSeek	29	685B 37B active at inference time	128k	$0.3	-	🤗		View
DeepSeek V3.2 Exp (Non-reasoning)	DeepSeek	28	685B 37B active at inference time	128k	$0.2	-	🤗		View
Nemotron Cascade 2 30B A3B	NVIDIA	28	31.6B 3B active at inference time	1.00M	-	-	🤗	-	View
Apriel-v1.5-15B-Thinker	ServiceNow	28	15B	128k	-	-	🤗		View
Qwen3 Coder Next	Alibaba	28	79.7B 3B active at inference time	256k	$0.4	113	🤗		View
DeepSeek V3.1 (Non-reasoning)	DeepSeek	28	685B 37B active at inference time	128k	$0.7	-	🤗	+7	View
Mistral Small 4 (Reasoning)	Mistral	28	119B 6.5B active at inference time	256k	$0.2	174	🤗		View
DeepSeek V3.1 (Reasoning)	DeepSeek	28	685B 37B active at inference time	128k	$0.7	-	🤗		View
Qwen3 VL 235B A22B (Reasoning)	Alibaba	28	235B 22B active at inference time	262k	$1.4	39	🤗		View
Apriel-v1.6-15B-Thinker	ServiceNow	28	15B	128k	-	-	🤗		View
Qwen3.5 9B (Non-reasoning)	Alibaba	27	9.65B	262k	-	-	🤗	-	View
Gemma 4 26B A4B (Non-reasoning)	Google	27	25.2B 3.8B active at inference time	256k	$0.2	60	🤗	+4	View
Qwen3.5 4B (Reasoning)	Alibaba	27	4.66B	262k	$0.0	200	🤗		View
DeepSeek R1 0528 (May '25)	DeepSeek	27	685B 37B active at inference time	128k	$1.6	-	🤗	+3	View
Qwen3 Next 80B A3B (Reasoning)	Alibaba	27	80B 3B active at inference time	262k	$1.1	169	🤗	+5	View
GLM-4.5 (Reasoning)	Z AI	26	355B 32B active at inference time	128k	$0.8	52	🤗		View
Kimi K2	Kimi	26	1.0KB 32B active at inference time	128k	$0.6	24	🤗		View
Ling 2.6 Flash	InclusionAI	26	107B 7.4B active at inference time	262k	$0.1	-	🤗		View
Seed-OSS-36B-Instruct	ByteDance Seed	25	36.2B	512k	$0.2	36	🤗		View
Qwen3 235B A22B 2507 Instruct	Alibaba	25	235B 22B active at inference time	256k	$0.3	53	🤗	+9	View
Qwen3 Coder 480B A35B Instruct	Alibaba	25	480B 35B active at inference time	262k	$0.5	59	🤗	+6	View
Qwen3 VL 32B (Reasoning)	Alibaba	25	33.4B	256k	$1.5	90	🤗		View
gpt-oss-20B (high)	OpenAI	24	21B 3.6B active at inference time	131k	$0.1	240	🤗	+10	View
gpt-oss-120b (low)	OpenAI	24	117B 5.1B active at inference time	131k	$0.2	357	🤗	+19	View
MiniMax M1 80k	MiniMax	24	456B 45.9B active at inference time	1.00M	$0.7	-	🤗		View
NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)	NVIDIA	24	31.6B 3.6B active at inference time	1.00M	$0.1	149	🤗		View
K2 Think V2	MBZUAI Institute of Foundation Models	24	70B	262k	-	-	🤗	-	View
LongCat Flash Lite	LongCat	24	68.5B 3B active at inference time	256k	-	-	🤗		View
HyperCLOVA X SEED Think (32B)	Naver	24	32B	128k	-	-	🤗	-	View
GLM-4.6V (Reasoning)	Z AI	23	108B 12B active at inference time	128k	$0.4	90	🤗		View
K-EXAONE (Non-reasoning)	LG AI Research	23	236B 23B active at inference time	256k	-	-	🤗	-	View
GLM-4.5-Air	Z AI	23	106B 12B active at inference time	128k	$0.3	80	🤗		View
Mistral Large 3	Mistral	23	675B 41B active at inference time	256k	$0.6	53	🤗		View
Ring-1T	InclusionAI	23	1.0KB 50B active at inference time	128k	-	-	🤗	-	View
Qwen3.5 4B (Non-reasoning)	Alibaba	23	4.66B	262k	$0.0	208	🤗		View
Qwen3 30B A3B 2507 (Reasoning)	Alibaba	22	30.5B 3.3B active at inference time	262k	$0.4	131	🤗		View
DeepSeek V3 0324	DeepSeek	22	671B 37B active at inference time	128k	$1.2	-	🤗	+3	View
INTELLECT-3	Prime Intellect	22	107B 12B active at inference time	131k	-	-	🤗	-	View
GLM-4.7-Flash (Non-reasoning)	Z AI	22	31.2B 3B active at inference time	200k	$0.1	102	🤗		View
Devstral 2	Mistral	22	125B	256k	-	61	🤗		View
Solar Open 100B (Reasoning)	Upstage	22	102B 12B active at inference time	128k	-	-	🤗	-	View
Nemotron 3 Nano Omni 30B A3B Reasoning	NVIDIA	21	30B 3B active at inference time	256k	$0.1	296	🤗		View
MiniMax M1 40k	MiniMax	21	456B 45.9B active at inference time	1.00M	-	-	🤗	-	View
gpt-oss-20B (low)	OpenAI	21	21B 3.6B active at inference time	131k	$0.1	243	🤗	+9	View
Qwen3 VL 235B A22B Instruct	Alibaba	21	235B 22B active at inference time	262k	$0.5	51	🤗	+2	View
K2-V2 (high)	MBZUAI Institute of Foundation Models	21	70B	512k	-	-	🤗	-	View
Qwen3 Next 80B A3B Instruct	Alibaba	20	80B 3B active at inference time	262k	$0.7	149	🤗	+4	View
Tri-21B-think Preview	Trillion Labs	20	21B	32.0k	-	-	🤗	-	View
Qwen3 Coder 30B A3B Instruct	Alibaba	20	30.5B 3.3B active at inference time	262k	$0.3	98	🤗		View
Qwen3 235B A22B (Reasoning)	Alibaba	20	235B 22B active at inference time	32.8k	$1.5	58	🤗		View
QwQ 32B	Alibaba	20	32.8B	131k	$0.7	30	🤗		View
Qwen3 VL 30B A3B (Reasoning)	Alibaba	20	30B 3B active at inference time	256k	$0.3	111	🤗		View
Devstral Small 2	Mistral	19	24B	256k	-	51	🤗		View
Ling-1T	InclusionAI	19	1.0KB 50B active at inference time	128k	-	-	🤗	-	View
DeepSeek R1 (Jan '25)	DeepSeek	19	685B 37B active at inference time	128k	$2.0	-	🤗	+3	View
Gemma 4 E4B (Reasoning)	Google	19	8B 4.5B active at inference time	128k	-	-	🤗	-	View
K2-V2 (medium)	MBZUAI Institute of Foundation Models	19	70B	512k	-	-	🤗	-	View
Llama Nemotron Super 49B v1.5 (Reasoning)	NVIDIA	19	49B	128k	$0.1	48	🤗		View
Mistral Small 4 (Non-reasoning)	Mistral	19	119B 6.5B active at inference time	256k	$0.2	164	🤗		View
Tri-21B-Think	Trillion Labs	19	21B	32.0k	-	-	🤗	-	View
Hermes 4 - Llama-3.1 405B (Reasoning)	Nous Research	19	406B	128k	$1.2	40	🤗		View
Llama 3.3 Nemotron Super 49B v1 (Reasoning)	NVIDIA	18	49B	128k	-	-	🤗	-	View
Llama 4 Maverick	Meta	18	402B 17B active at inference time	1.00M	$0.3	111	🤗	+6	View
Qwen3 4B 2507 (Reasoning)	Alibaba	18	4.02B	262k	-	-	🤗	-	View
MiniCPM5-1B (Reasoning)	OpenBMB	18	1B	128k	-	-	🤗	-	View
Magistral Small 1.2	Mistral	18	24B	128k	$0.6	109	🤗		View
Sarvam 105B (high)	Sarvam	18	106B 10.3B active at inference time	128k	$0.0	109	🤗		View
Devstral Small (May '25)	Mistral	18	23.6B	256k	-	-	🤗	-	View
MiniCPM5-1B (Non-reasoning)	OpenBMB	18	1B	128k	-	-	🤗	-	View
Hermes 4 - Llama-3.1 405B (Non-reasoning)	Nous Research	18	406B	128k	$1.2	41	🤗		View
Llama 3.1 Instruct 405B	Meta	17	405B	128k	$3.1	65	🤗		View
Qwen3 VL 32B Instruct	Alibaba	17	33.4B	256k	$0.9	72	🤗		View
DeepSeek R1 Distill Qwen 32B	DeepSeek	17	32B	128k	-	-	🤗	-	View
GLM-4.6V (Non-reasoning)	Z AI	17	108B 12B active at inference time	128k	$0.4	93	🤗		View
Qwen3 235B A22B (Non-reasoning)	Alibaba	17	235B 22B active at inference time	32.8k	$0.6	57	🤗		View
Magistral Small 1	Mistral	17	23.6B	40.0k	-	-	🤗	-	View
EXAONE 4.0 32B (Reasoning)	LG AI Research	17	32B	131k	-	-	🤗	-	View
Qwen3 VL 8B (Reasoning)	Alibaba	17	8.77B	256k	$0.4	113	🤗		View
Qwen3 32B (Reasoning)	Alibaba	17	32.8B	32.8k	$0.2	81	🤗	+3	View
DeepSeek V3 (Dec '24)	DeepSeek	16	671B 37B active at inference time	128k	$0.4	-	🤗	+2	View
DeepSeek R1 0528 Qwen3 8B	DeepSeek	16	8.19B	32.8k	-	-	🤗	-	View
Qwen3.5 2B (Reasoning)	Alibaba	16	2.27B	262k	$0.0	-	🤗		View
Qwen3 14B (Reasoning)	Alibaba	16	14.8B	32.8k	$0.4	63	🤗		View
Nanbeige4.1-3B	Nanbeige	16	3.93B	256k	-	-	🤗	-	View
Qwen3 VL 30B A3B Instruct	Alibaba	16	30B 3B active at inference time	256k	$0.2	104	🤗		View
Hermes 4 - Llama-3.1 70B (Reasoning)	Nous Research	16	70.6B	128k	$0.2	88	🤗		View
Ministral 3 14B	Mistral	16	14B	256k	$0.2	81	🤗		View
DeepSeek R1 Distill Llama 70B	DeepSeek	16	70B	128k	$0.7	41	🤗		View
DeepSeek R1 Distill Qwen 14B	DeepSeek	16	14B	128k	-	-	🤗	-	View
Falcon-H1R-7B	TII UAE	16	7B	256k	-	-	🤗	-	View
Ling-flash-2.0	InclusionAI	16	103B 6.1B active at inference time	128k	$0.2	80	🤗		View
Qwen3 Omni 30B A3B (Reasoning)	Alibaba	16	35.3B 3B active at inference time	65.5k	$0.3	86	🤗		View
Qwen2.5 Instruct 72B	Alibaba	16	72B	131k	$0.2	-	🤗		View
Step3 VL 10B	StepFun	15	10.2B	65.5k	-	-	🤗	-	View
Qwen3 30B A3B (Reasoning)	Alibaba	15	30.5B 3.3B active at inference time	32.8k	$0.1	75	🤗	+2	View
Devstral Small (Jul '25)	Mistral	15	24B	256k	$0.1	42	🤗		View
Gemma 4 E2B (Reasoning)	Google	15	5.1B 2.3B active at inference time	128k	-	-	🤗	-	View
QwQ 32B-Preview	Alibaba	15	32.8B	32.8k	-	-	🤗	-	View
GLM-4.5V (Reasoning)	Z AI	15	108B 12B active at inference time	64.0k	$0.7	26	🤗		View
Mistral Large 2 (Nov '24)	Mistral	15	123B	128k	$2.4	52	🤗		View
Mistral Small 3.2	Mistral	15	24B	128k	$0.1	139	🤗		View
Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)	NVIDIA	15	253B	128k	$0.7	52	🤗		View
Qwen3 30B A3B 2507 Instruct	Alibaba	15	30.5B 3.3B active at inference time	262k	$0.2	132	🤗		View
ERNIE 4.5 300B A47B	Baidu	15	300B 47B active at inference time	131k	$0.4	-	🤗		View
NVIDIA Nemotron Nano 12B v2 VL (Reasoning)	NVIDIA	15	13.2B	128k	$0.2	286	🤗		View
Ministral 3 8B	Mistral	15	8B	256k	$0.1	106	🤗		View
Gemma 4 E4B (Non-reasoning)	Google	15	8B 4.5B active at inference time	128k	-	-	🤗	-	View
NVIDIA Nemotron Nano 9B V2 (Reasoning)	NVIDIA	15	9B	131k	$0.1	122	🤗		View
Granite 4.1 30B	IBM	15	30B	131k	-	-	🤗	-	View
NVIDIA Nemotron 3 Nano 4B	NVIDIA	15	3.97B	262k	-	-	🤗	-	View
Qwen3.5 2B (Non-reasoning)	Alibaba	15	2.27B	262k	$0.0	344	🤗		View
Llama Nemotron Super 49B v1.5 (Non-reasoning)	NVIDIA	15	49B	128k	$0.1	49	🤗		View
Qwen3 32B (Non-reasoning)	Alibaba	15	32.8B	32.8k	$0.2	88	🤗	+4	View
Llama 3.3 Instruct 70B	Meta	14	70B	128k	$0.6	86	🤗	+18	View
Mistral Small 3.1	Mistral	14	24B	128k	$0.1	173	🤗		View
K2-V2 (low)	MBZUAI Institute of Foundation Models	14	70B	512k	-	-	🤗	-	View
Llama 3.1 Nemotron Nano 4B v1.1 (Reasoning)	NVIDIA	14	4.51B	128k	-	-	🤗	-	View
Kimi Linear 48B A3B Instruct	Kimi	14	49.1B 3B active at inference time	1.00M	-	-	🤗	-	View
Llama 3.3 Nemotron Super 49B v1 (Non-reasoning)	NVIDIA	14	49B	128k	-	-	🤗	-	View
Qwen3 VL 8B Instruct	Alibaba	14	8.77B	256k	$0.2	123	🤗		View
Qwen3 4B (Reasoning)	Alibaba	14	4.02B	32.0k	$0.2	-	🤗		View
Llama 3.1 Tulu3 405B	Allen Institute for AI	14	405B	128k	-	-	🤗	-	View
Ring-flash-2.0	InclusionAI	14	103B 6.1B active at inference time	128k	$0.2	-	🤗		View
Pixtral Large	Mistral	14	124B	128k	$2.4	52	🤗		View
Olmo 3.1 32B Think	Allen Institute for AI	14	32.2B	65.5k	-	-	🤗		View
Grok 2 (Dec '24)	xAI	14	270B	131k	-	-	🤗	-	View
Qwen3 VL 4B (Reasoning)	Alibaba	14	4.44B	256k	-	-	🤗	-	View
Llama 4 Scout	Meta	14	109B 17B active at inference time	10.0M	$0.2	101	🤗	+6	View
Command A	Cohere	13	111B	256k	$3.3	65	🤗		View
Llama 3.1 Nemotron Instruct 70B	NVIDIA	13	70B	128k	$1.2	300	🤗		View
Qwen2.5 Instruct 32B	Alibaba	13	32B	128k	-	-	🤗	-	View
Qwen3 8B (Reasoning)	Alibaba	13	8.19B	131k	$0.2	40	🤗		View
NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)	NVIDIA	13	31.6B 3.6B active at inference time	1.00M	$0.1	93	🤗		View
NVIDIA Nemotron Nano 9B V2 (Non-reasoning)	NVIDIA	13	9B	131k	$0.1	130	🤗		View
Mistral Large 2 (Jul '24)	Mistral	13	123B	128k	$2.4	-	🤗		View
Qwen3 4B 2507 Instruct	Alibaba	13	4.02B	262k	-	-	🤗	-	View
Qwen2.5 Coder Instruct 32B	Alibaba	13	32B	131k	-	-	🤗	-	View
Qwen3 14B (Non-reasoning)	Alibaba	13	14.8B	32.8k	$0.3	62	🤗		View
GLM-4.5V (Non-reasoning)	Z AI	13	108B 12B active at inference time	64.0k	$0.7	38	🤗		View
Mistral Small 3	Mistral	13	24B	32.0k	$0.1	167	🤗		View
MiniCPM-V 4.6 1.3B	OpenBMB	13	1.3B	262k	-	-	🤗	-	View
Hermes 4 - Llama-3.1 70B (Non-reasoning)	Nous Research	13	70.6B	128k	$0.2	88	🤗		View
Qwen3 30B A3B (Non-reasoning)	Alibaba	13	30.5B 3.3B active at inference time	32.8k	$0.1	76	🤗		View
DeepSeek-V2.5 (Dec '24)	DeepSeek	13	236B 21B active at inference time	128k	-	-	🤗	-	View
Qwen3 4B (Non-reasoning)	Alibaba	12	4.02B	32.0k	$0.1	-	🤗		View
Llama 3.1 Instruct 70B	Meta	12	70B	128k	$0.6	34	🤗		View
Granite 4.1 8B	IBM	12	8B	131k	$0.1	122	🤗		View
Sarvam 30B (high)	Sarvam	12	32.2B 2.4B active at inference time	65.5k	$0.0	166	🤗		View
DeepSeek-V2.5	DeepSeek	12	236B 21B active at inference time	128k	-	-	🤗	-	View
Olmo 3.1 32B Instruct	Allen Institute for AI	12	32.2B	65.5k	-	-	🤗	-	View
DeepSeek R1 Distill Llama 8B	DeepSeek	12	8B	128k	-	-	🤗	-	View
Gemma 4 E2B (Non-reasoning)	Google	12	5.1B 2.3B active at inference time	128k	-	-	🤗	-	View
Olmo 3 32B Think	Allen Institute for AI	12	32.2B	65.5k	-	-	🤗	-	View
R1 1776	Perplexity	12	671B 37B active at inference time	128k	-	-	🤗	-	View
Llama 3.2 Instruct 90B (Vision)	Meta	12	90B	128k	$1.4	59	🤗		View
Solar Mini	Upstage	12	10.7B	4.10k	$0.1	-	🤗		View
Llama 3.1 Instruct 8B	Meta	12	8B	128k	$0.1	155	🤗	+12	View
Grok-1	xAI	12	314B 78B active at inference time	8.19k	-	-	🤗	-	View
Qwen2 Instruct 72B	Alibaba	12	72B	131k	-	-	🤗	-	View
EXAONE 4.0 32B (Non-reasoning)	LG AI Research	12	32B	131k	-	-	🤗	-	View
Ministral 3 3B	Mistral	11	3B	256k	$0.1	176	🤗		View
DeepHermes 3 - Mistral 24B Preview (Non-reasoning)	Nous Research	11	24B	32.0k	-	-	🤗	-	View
Jamba 1.7 Large	AI21 Labs	11	398B 94B active at inference time	256k	$2.6	60	🤗		View
Granite 4.0 H Small	IBM	11	32B 9B active at inference time	128k	$0.1	357	🤗		View
Jamba 1.5 Large	AI21 Labs	11	398B 94B active at inference time	256k	$2.6	-	🤗		View
Qwen3 Omni 30B A3B Instruct	Alibaba	11	35.3B 3B active at inference time	65.5k	$0.3	94	🤗		View
Hermes 3 - Llama-3.1 70B	Nous Research	11	70.6B	128k	$0.3	29	🤗		View
Qwen3 8B (Non-reasoning)	Alibaba	11	8.19B	32.8k	$0.2	40	🤗		View
DeepSeek-Coder-V2	DeepSeek	11	236B 21B active at inference time	128k	-	-	🤗	-	View
OLMo 2 32B	Allen Institute for AI	11	32.2B	4.10k	-	-	🤗	-	View
Jamba 1.6 Large	AI21 Labs	11	398B 94B active at inference time	256k	$2.6	61	🤗		View
Qwen3.5 0.8B (Reasoning)	Alibaba	11	0.873B	262k	$0.0	-	🤗		View
LFM2 24B A2B	Liquid AI	10	23.8B 2.3B active at inference time	32.8k	$0.0	118	🤗		View
Phi-4	Microsoft	10	14B	16.0k	$0.2	33	🤗		View
Gemma 3 27B Instruct	Google	10	27.4B	128k	$0.1	-	🤗	+3	View
Mistral Small (Sep '24)	Mistral	10	22B	32.8k	$0.2	172	🤗		View
Phi-3 Mini Instruct 3.8B	Microsoft	10	3.8B	4.10k	-	-	🤗	-	View
NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning)	NVIDIA	10	13.2B	128k	$0.2	221	🤗		View
Gemma 3n E4B Instruct Preview (May '25)	Google	10	8.39B 4B active at inference time	32.0k	-	-	🤗	-	View
Phi-4 Multimodal Instruct	Microsoft	10	5.6B	128k	-	17	🤗		View
Qwen2.5 Coder Instruct 7B	Alibaba	10	7.62B	131k	-	-	🤗	-	View
Qwen3.5 0.8B (Non-reasoning)	Alibaba	10	0.873B	262k	$0.0	91	🤗		View
Mixtral 8x22B Instruct	Mistral	10	141B 39B active at inference time	65.4k	-	-	🤗	-	View
Llama 2 Chat 7B	Meta	10	7B	4.10k	$0.1	-	🤗		View
Llama 3.2 Instruct 3B	Meta	10	3B	128k	$0.1	51	🤗		View
Jamba Reasoning 3B	AI21 Labs	10	3B	262k	-	-	🤗	-	View
Qwen3 VL 4B Instruct	Alibaba	10	4.44B	256k	-	-	🤗	-	View
Qwen1.5 Chat 110B	Alibaba	10	110B	32.0k	-	-	🤗	-	View
Reka Flash 3	Reka AI	10	21B	128k	$0.3	-	🤗		View
Olmo 3 7B Think	Allen Institute for AI	9	7B	65.5k	-	-	🤗	-	View
OLMo 2 7B	Allen Institute for AI	9	7.3B	4.10k	-	-	🤗	-	View
Molmo 7B-D	Allen Institute for AI	9	8.02B	4.10k	-	-	🤗	-	View
Ling-mini-2.0	InclusionAI	9	16.3B 1.4B active at inference time	131k	-	-	🤗	-	View
DeepSeek R1 Distill Qwen 1.5B	DeepSeek	9	1.5B	128k	-	-	🤗	-	View
DeepSeek-V2-Chat	DeepSeek	9	236B 21B active at inference time	128k	-	-	🤗	-	View
Llama 3 Instruct 70B	Meta	9	70B	8.19k	$0.9	-	🤗		View
Arctic Instruct	Snowflake	9	480B 17B active at inference time	4.00k	-	-	🤗	-	View
Qwen Chat 72B	Alibaba	9	72B	33.8k	-	-	🤗	-	View
Gemma 3 12B Instruct	Google	9	12.2B	128k	$0.1	-	🤗	+2	View
Llama 3.2 Instruct 11B (Vision)	Meta	9	11B	128k	$0.2	51	🤗		View
Granite 4.1 3B	IBM	9	3B	131k	-	-	🤗	-	View
DeepSeek Coder V2 Lite Instruct	DeepSeek	8	16B 2.4B active at inference time	128k	-	-	🤗	-	View
Sarvam M (Reasoning)	Sarvam	8	23.6B	32.8k	-	-	🤗		View
Phi-4 Mini Instruct	Microsoft	8	3.84B	128k	-	22	🤗		View
Llama 2 Chat 70B	Meta	8	70B	4.10k	-	-	🤗	-	View
DeepSeek LLM 67B Chat (V1)	DeepSeek	8	7B	4.10k	-	-	🤗	-	View
Llama 2 Chat 13B	Meta	8	13B	4.10k	-	-	🤗	-	View
Command-R+ (Apr '24)	Cohere	8	104B	128k	$4.2	-	🤗		View
OpenChat 3.5 (1210)	OpenChat	8	7B	8.19k	-	-	🤗	-	View
DBRX Instruct	Databricks	8	132B 36B active at inference time	32.8k	-	-	🤗	-	View
Exaone 4.0 1.2B (Reasoning)	LG AI Research	8	1.28B	64.0k	-	-	🤗	-	View
Olmo 3 7B Instruct	Allen Institute for AI	8	7B	65.5k	$0.1	-	🤗		View
Exaone 4.0 1.2B (Non-reasoning)	LG AI Research	8	1.28B	64.0k	-	-	🤗	-	View
LFM2.5-1.2B-Thinking	Liquid AI	8	1.17B	32.0k	-	-	🤗	-	View
Jamba 1.7 Mini	AI21 Labs	8	52B 12B active at inference time	258k	-	-	🤗	-	View
LFM2 2.6B	Liquid AI	8	2.57B	32.8k	-	-	🤗	?	View
LFM2.5-1.2B-Instruct	Liquid AI	8	1.17B	32.0k	-	-	🤗	?	View
Jamba 1.5 Mini	AI21 Labs	8	52B 12B active at inference time	256k	$0.2	-	🤗		View
Granite 4.0 H 1B	IBM	8	1.5B	128k	-	-	🤗	-	View
Qwen3 1.7B (Reasoning)	Alibaba	8	2.03B	32.0k	$0.2	-	🤗		View
Jamba 1.6 Mini	AI21 Labs	8	52B 12B active at inference time	256k	$0.2	184	🤗		View
Mixtral 8x7B Instruct	Mistral	8	46.7B 12.9B active at inference time	32.8k	$0.5	-	🤗		View
Gemma 3 270M	Google	8	0.268B	32.0k	-	-	🤗	-	View
Apertus 70B Instruct	Swiss AI Initiative	8	70B	65.5k	$1.0	-	🤗		View
Granite 4.0 Micro	IBM	8	3B	128k	-	-	🤗	-	View
DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning)	Nous Research	8	8B	128k	-	-	🤗	-	View
Llama 65B	Meta	7	65B	2.05k	-	-	🤗	-	View
Qwen Chat 14B	Alibaba	7	14B	8.19k	-	-	🤗	-	View
Mistral 7B Instruct	Mistral	7	7B	8.19k	$0.2	98	🤗		View
Command-R (Mar '24)	Cohere	7	35B	128k	$0.6	-	🤗		View
Granite 4.0 1B	IBM	7	1.6B	128k	-	-	🤗	-	View
Molmo2-8B	Allen Institute for AI	7	8.66B	36.9k	-	-	🤗	-	View
LFM2 8B A1B	Liquid AI	7	8.34B 1.5B active at inference time	32.8k	-	-	🤗	?	View
Granite 3.3 8B (Non-reasoning)	IBM	7	8.17B	128k	$0.1	408	🤗		View
Qwen3 1.7B (Non-reasoning)	Alibaba	7	2.03B	32.0k	$0.1	-	🤗		View
Qwen3 0.6B (Reasoning)	Alibaba	6	0.752B	32.0k	$0.2	-	🤗		View
Llama 3 Instruct 8B	Meta	6	8B	8.19k	$0.1	-	🤗		View
Gemma 3n E4B Instruct	Google	6	8.39B 4B active at inference time	32.0k	$0.0	53	🤗		View
LFM2 1.2B	Liquid AI	6	1.17B	32.8k	-	-	🤗	?	View
Gemma 3 4B Instruct	Google	6	4.3B	128k	$0.0	-	🤗		View
Llama 3.2 Instruct 1B	Meta	6	1B	128k	$0.1	84	🤗		View
LFM2.5-VL-1.6B	Liquid AI	6	1.6B	32.0k	-	-	🤗	?	View
Granite 4.0 350M	IBM	6	0.35B	32.8k	-	-	🤗	-	View
Apertus 8B Instruct	Swiss AI Initiative	6	8B	65.5k	$0.1	-	🤗		View
Qwen3 0.6B (Non-reasoning)	Alibaba	6	0.752B	32.0k	$0.1	-	🤗		View
Gemma 3 1B Instruct	Google	6	1B	32.0k	-	-	🤗		View
Granite 4.0 H 350M	IBM	5	0.34B	32.8k	-	-	🤗	-	View
Gemma 3n E2B Instruct	Google	5	5.98B 2B active at inference time	32.0k	-	-	🤗		View
Tiny Aya Global	Cohere	5	3.35B	8.19k	-	-	🤗		View
EXAONE 4.5 33B (Non-reasoning)	LG AI Research	-	34.4B	262k	-	-	🤗	-	View
Cogito v2.1 (Reasoning)	Deep Cogito	-	671B 37B active at inference time	128k	$1.3	69	🤗		View

Comparison of Open Source Models

Openness

Intelligence

Total Parameters

Openness

Artificial Analysis Openness Index: Results

Open Source Progress

Progress in Open Weights vs. Proprietary Intelligence

Artificial Analysis Intelligence Index

Open Weights

Open Source Language Models Intelligence By Lab Over Time

Artificial Analysis Intelligence Index

Open Source Models Intelligence By Size Over Time

Artificial Analysis Intelligence Index

Model Size Classifications

Intelligence

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index

Intelligence Evaluations

Intelligence Evaluation Relevance

Artificial Analysis Intelligence Index

Size

Intelligence Index By Model Size

Artificial Analysis Intelligence Index

Open Weights

Model Size Classifications

Model Size: Total and Active Parameters

Total Parameters

Active Parameters at Inference Time

Intelligence vs. Active Parameters

Artificial Analysis Intelligence Index

Active Parameters at Inference Time

Intelligence vs. Total Parameters

Artificial Analysis Intelligence Index

Total Parameters

Context Window

Context Window

Context Window for RAG

Context Window

Further details