llmboard.aiLeaderboard Center
Overall
Overall RankingOpen Models
Tools
Model DirectoryCompare Models
Capabilities
CodingReasoningMathKnowledgeInstruction Following
Price & Efficiency
Price & ValueCapability vs. PriceRuntime Performance
Modalities
Image GenerationVideo GenerationSpeech ModelsEmbeddings
Core Benchmarks
GPQAMMLU-ProAIME 2025SWE-Bench VerifiedMMLUHumanity's Last ExamLiveCodeBenchMATHHumanEvalMMMU-ProView all benchmarks
Methods
Scoring & Data
393 models668 benchmarks

Leaderboard Center

Overall RankingCodingCore BenchmarksPrice & ValueRuntime Performance

Modalities

All ModelsImage GenerationImage EditingVideo GenerationImage-to-VideoVideo EditingText-to-SpeechSpeech-to-TextEmbeddings

Data & Methods

Scoring MethodAll BenchmarksReasoningMath

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai

Open weights

Open Model Leaderboard

Best Open-Weight LLMs

Data as of 2026-08-11

On this page

  • Overview
  • Capability
  • Landscape
  • Ranking
  • Value
  • Compare
  • FAQ

Open Model Ranking Overview

Top open-weight modelKimi K393.0 LLMBoard scoreLargest contextKimi K31M tokensMost recent releaseMuse Glimmer 30B2026-08-10
Ranked open models60Unique model products

Open Model Capability Leaders

Each capability remains a separate score.

Reasoning97.4Kimi K3Rank #1 overall
Knowledge91.3Kimi K3Rank #1 overall
Math100.0MiniMax M3Rank #5 overall
Coding96.9Llama 3.3 Nemotron Super 49BRank #50 overall
Instruction90.9Llama 3.3 Nemotron Super 49BRank #50 overall

Open-model landscape

Vendor, license and model-size patterns across the ranked models.

Top-30 vendor mix

30 models
Zhipu AI6Alibaba Cloud / Qwen Team6Moonshot AI4MiniMax4Xiaomi2Google2DeepSeek1Tencent1

Scored-version licenses

Apache 2.0
27
MIT
20
Modified MIT License
3
NVIDIA Open Model License Agreement
3
Llama 3.1 Community License
2
Kimi K3 License
1
OpenMDW License v1.1
1
02550751002.3B15.3B103.7B701.4B4.7TLLMBoard scoreParameters
59 models with comparable dataOpen a model by selecting its logo

Open Model Ranking

Each model is ranked using its latest scored open-weight version.

60 rows
Columns

Show columns

01MAKimi K3Moonshot AI93.01MOpen weightsJul 16, 2026Kimi K3
02ZAGLM 5.2Zhipu AI79.31MOpen weightsJun 16, 2026GLM-5.2
03DEDeepSeek-V4-FlashDeepSeek78.31MOpen weightsJul 31, 2026DeepSeek-V4-Flash-0731
04MAKimi K2.6Moonshot AI75.5262.1KOpen weightsApr 20, 2026Kimi K2.6
05MIMiniMax M3MiniMax71.3512KOpen weightsJun 1, 2026MiniMax M3
06TEHy3Tencent71.0256KOpen weightsJul 6, 2026Hy3
07ZAGLM 5.1Zhipu AI69.6200KOpen weightsApr 7, 2026GLM-5.1
08ZAGLM 5Zhipu AI67.2204.8KOpen weightsFeb 11, 2026GLM-5
09MAKimi K2.7 CodeMoonshot AI66.8262.1KOpen weightsJun 12, 2026Kimi K2.7 Code
10ACQwen3.5 397B A17BAlibaba Cloud / Qwen Team66.3262.1KOpen weightsFeb 16, 2026Qwen3.5-397B-A17B
11MAKimi K2.5Moonshot AI66.0262.1KOpen weightsJan 27, 2026Kimi K2.5
12NVNemotron 3 UltraNVIDIA62.91MOpen weightsJun 4, 2026Nemotron 3 Ultra (550B A55B)
13XIMiMoXiaomi61.81MOpen weightsApr 22, 2026MiMo-V2.5
14ACQwen3.6 27BAlibaba Cloud / Qwen Team60.5262.1KOpen weightsApr 21, 2026Qwen3.6-27B
15STStep 3.5 FlashStepFun60.4256KOpen weightsFeb 2, 2026Step-3.5-Flash
16GOGemma 4 31BGoogle60.1262.1KOpen weightsApr 2, 2026Gemma 4 31B
17ZAGLM 4.7Zhipu AI59.7204.8KOpen weightsDec 22, 2025GLM-4.7
18ACQwen3.5 122B A10BAlibaba Cloud / Qwen Team59.2262.1KOpen weightsFeb 24, 2026Qwen3.5-122B-A10B
19MEMuse Glimmer 30BMeta58.5131.1KOpen weightsAug 10, 2026Muse Glimmer-30B
20MIMiniMax M2.7MiniMax58.4204.8KOpen weightsMar 18, 2026MiniMax M2.7
21ACQwen3.5 27BAlibaba Cloud / Qwen Team57.2262.1KOpen weightsFeb 24, 2026Qwen3.5-27B
22MIMiniMax M2.5MiniMax55.6204.8KOpen weightsFeb 12, 2026MiniMax M2.5
23ACQwen3.6 35B A3BAlibaba Cloud / Qwen Team55.3262.1KOpen weightsApr 16, 2026Qwen3.6-35B-A3B
24ZAGLM 4.6Zhipu AI54.2204.8KOpen weightsSep 30, 2025GLM-4.6
25MIMiniMax M2.1MiniMax53.7204.8KOpen weightsDec 23, 2025MiniMax M2.1
26GOGemma 4 26B A4BGoogle52.6262.1KOpen weightsApr 2, 2026Gemma 4 26B-A4B
27ACQwen3.5 35B A3BAlibaba Cloud / Qwen Team52.0262.1KOpen weightsFeb 24, 2026Qwen3.5-35B-A3B
28XIMiMo V2 FlashXiaomi51.3262.1KOpen weightsDec 16, 2025MiMo-V2-Flash
29ZAGLM 4.5Zhipu AI49.3131.1KOpen weightsJul 28, 2025GLM-4.5
30MAMistral Medium 3.5Mistral AI49.2262.1KOpen weightsApr 29, 2026Mistral Medium 3.5
31XIMiMo V2.5 ProXiaomi46.71MOpen weightsApr 27, 2026MiMo-V2.5-Pro
32NVNemotron 3 SuperNVIDIA44.8262.1KOpen weightsMar 11, 2026Nemotron 3 Super (120B A12B)
33SASarvam 105BSarvam AI44.6131.1KOpen weightsMar 6, 2026Sarvam-105B
34MIMiniMax M2MiniMax43.9196.6KOpen weightsOct 27, 2025MiniMax M2
35ACQwen3 235B A22BAlibaba Cloud / Qwen Team42.4262.1KOpen weightsJul 22, 2025Qwen3-235B-A22B-Instruct-2507
36COCommand A+Cohere41.9128KOpen weightsMay 20, 2026Command A+
37ZAGLM 4.5 AirZhipu AI41.5131.1KOpen weightsJul 28, 2025GLM-4.5-Air
38ACQwen3.5 9BAlibaba Cloud / Qwen Team41.4262.1KOpen weightsMar 2, 2026Qwen3.5-9B
39ACQwen3 Next 80B A3B ThinkingAlibaba Cloud / Qwen Team40.9131.1KOpen weightsSep 10, 2025Qwen3-Next-80B-A3B-Thinking
40DEDeepSeek-V3.1DeepSeek39.9131.1KOpen weightsJan 10, 2025DeepSeek-V3.1
41ZAGLM 4.7 FlashZhipu AI38.5200KOpen weightsJan 19, 2026GLM-4.7-Flash
42ACQwen3 Next 80B A3BAlibaba Cloud / Qwen Team36.6131.1KOpen weightsSep 10, 2025Qwen3-Next-80B-A3B-Instruct
43ACQwen3 Coder 480B A35BAlibaba Cloud / Qwen Team35.5262.1KOpen weightsJan 31, 2025Qwen3-Coder 480B A35B Instruct
44NVNemotron 3 NanoNVIDIA35.2262.1KOpen weightsDec 15, 2025Nemotron 3 Nano (30B A3B)
45SASarvam 30BSarvam AI34.3128KOpen weightsMar 6, 2026Sarvam-30B
46MAMistral Small 4Mistral AI33.4256KOpen weightsMar 16, 2026Mistral Small 4
47NVNemotron Nano 9BNVIDIA32.4131.1KOpen weightsAug 18, 2025Nemotron Nano 9B v2
48ACQwen3 32BAlibaba Cloud / Qwen Team31.2131.1KOpen weightsApr 29, 2025Qwen3 32B
49GOGemma 4 E4BGoogle24.4131.1KOpen weightsApr 2, 2026Gemma 4 E4B
50NVLlama 3.3 Nemotron Super 49BNVIDIA23.0131.1KOpen weightsMar 18, 2025Llama-3.3 Nemotron Super 49B v1
51ACQwen2.5 VL 72BAlibaba Cloud / Qwen Team22.7131.1KOpen weightsJan 26, 2025Qwen2.5 VL 72B Instruct
52MELlama 3.3 70BMeta22.1128KOpen weightsDec 6, 2024Llama 3.3 70B Instruct
53MAMagistral SmallMistral AI19.8131.1KOpen weightsJun 10, 2025Magistral Small 2506
54MADevstral Small 1.1Mistral AI17.1128KOpen weightsJul 11, 2025Devstral Small 1.1
55MAMistral Large 3Mistral AI15.8262.1KOpen weightsDec 4, 2025Mistral Large 3 (675B Instruct 2512)
56GOGemma 4 E2BGoogle11.6131.1KOpen weightsApr 2, 2026Gemma 4 E2B
57NVLlama 3.1 Nemotron 70BNVIDIA3.1128KOpen weightsOct 1, 2024Llama 3.1 Nemotron 70B Instruct
58MAPixtral 12BMistral AI0.0128KOpen weightsSep 17, 2024Pixtral-12B
59MAMinistral 8BMistral AI0.0131.1KOpen weightsOct 16, 2024Ministral 8B Instruct
60MIPhi 4 MiniMicrosoft0.0128KOpen weightsFeb 1, 2025Phi 4 Mini

Open-weight status belongs to the scored model version. License terms still need to be checked on the model detail page.

Official price and capability

Only positive official vendor PAYG prices are included.

020406080100$0.1$0.2$0.5$1.0$2.0$5.0LLMBoard scoreOfficial API price blend (8:1 input/output), USD / 1M
Efficient frontier41 models with official PAYG prices

Compare leading open models

#2-13.7
ZA

GLM 5.2

Zhipu AI

79.3 LLMBoard

DetailsCompare
#3-14.8
DE

DeepSeek-V4-Flash

DeepSeek

78.3 LLMBoard

DetailsCompare
#4-17.5
MA

Kimi K2.6

Moonshot AI

75.5 LLMBoard

DetailsCompare
#5-21.7
MI

MiniMax M3

MiniMax

71.3 LLMBoard

DetailsCompare
#6-22.1
TE

Hy3

Tencent

71.0 LLMBoard

DetailsCompare
#7-23.4
ZA

GLM 5.1

Zhipu AI

69.6 LLMBoard

DetailsCompare

About open-weight models

What is the highest-ranked open-weight model?

Kimi K3 currently leads the open-model ranking.

What does open weights mean here?

The scored model version is marked as having publicly available weights. That flag does not by itself determine whether the license permits every commercial or derivative use.

How many models are ranked?

This page currently ranks 60 unique open-weight models.

Which version is evaluated?

Each row names the latest open-weight version with a current LLMBoard score.