Skip to content
AI Atlas
Open model frontier

Open models

Every canonical model whose weights can be downloaded, described by measurable properties — licence permissions, parameters, context, release, benchmark ranks, providers, estimated local fit. Openness is a set of observed dimensions here, never a score.

1,062 downloadable models

Open sourceWeights and code published under an OSI-approved licence that allows commercial use and derivatives.

10

Open weightsWeights downloadable under a permissive or Creative Commons licence allowing commercial use; code or data may be missing.

970

Restricted weightsWeights downloadable, but the licence restricts commercial use, hosting, derivatives or field of use (community, research and RAIL licences).

0

Downloadable totalUniverse = canonical models whose weights can be downloaded (open-weights, open-source, restricted-weights). hardware_fit values are ESTIMATES (see /methodology); best_results are ranks inside each benchmark's primary comparability group — no composite score.

1,062

New in 30 daysDownloadable models with a release date inside the last 30 days (release_date claim; models without one are not counted).

264

Openness explorer

Categories and licence permissions

Left: the downloadable universe by openness category. Right: what the licences on this page allow, counted per category — allowed · restricted · unknown.

Top licences: Apache License 2.0 (358) · Other (unclassified licence) (157) · MIT License (105) · Creative Commons Attribution-NonCommercial 4.0 (22) · Apple Sample Code / ML Research License (15) · Llama 3.2 Community License (12)

Category · this pageCommercial useRedistributionDerivativesHosting
Open source n=0————
Open weights n=508042804280428042
restricted-weights n=0————

Counts read: allowed · restricted · unknown / unclassified — over the 50 models on this page, from each licence's stated terms (ontology).

Downloadable models

1,062 models · sorted by best benchmark rank

Permissions, in order:commercial use· redistribution· derivatives· hosting allowed restricted unknown

ModelLicence · permissionsParamsContextReleaseBest resultsProvidersFit 64 GB @4bit · 128 GB @8bit Estimated
gpt-oss-120bOpen weightsOpenAI · textApache-2.0116.8B131.1K4 Aug 2025swe-bench-verified#36ifbench#77scicode#1464from $0.17 out✗ too large67.7 GB4bit✗ too large134.8 GB8bit
command-a-plusOpen weightsCohere—unclassified—192K—ifbench#38tau2-bench#127scicode#1320——
kimi-k2-6-non-reasoningOpen weightsMoonshot AI—unclassified—256K—tau2-bench#39terminal-bench#63artificial-analysis-intelligence-index#1430——
gemma-4-12BOpen weightsGoogle—unclassified—256K—ifbench#40mmmu-pro#127terminal-bench#1810——
Qwen2.5 Coder 32B InstructOpen weightsQwen · text—unclassified—32.8K11 Nov 2024swe-bench-verified#42aider-polyglot#60artificial-analysis-intelligence-index#4891from $1 out——
qwen3-6-27b-non-reasoningOpen weightsAlibaba Group—unclassified—262.1K—tau2-bench#42mmmu-pro#111terminal-bench#1660——
Qwen3 32BOpen weightsQwen · textApache-2.032.8B131.1K27 Apr 2025aider-polyglot#44scicode#144tau2-bench#2751from $0.28 out✓ fits19.3 GB4bit✓ fits38.2 GB8bit
mimo-v2-0206Open weightsXiaomi—unclassified—256K—tau2-bench#44ifbench#52terminal-bench#1110——
tri-21b-think-previewOpen weightsTrillion Labs—unclassified—32K—tau2-bench#44ifbench#200terminal-bench#3330——
Ling 3.0 Flash VLOpen weightsinclusionAI · image, text, video—unclassified—131.1K10 Sept 2026mmmu-pro#47scicode#103gpqa#1201from $0.18 out——
Qwen3.5-35B-A3BOpen weightsQwen · image, text, video—unclassified—262.1K25 Feb 2026ifbench#48tau2-bench#73mmmu-pro#1061from $1.25 out——
Gemma 4 26B A4BOpen weightsGoogle · image, text, videoApache-2.025.8B4B active262.1K11 Mar 2026ifbench#49mmmu-pro#128humanitys-last-exam#1791from $0.22 out✓ fits15.3 GB4bit✓ fits30.2 GB8bit
MiMo-V2.5Open weightsXiaomi · audio, image, text, video—unclassified—1.05M22 Apr 2026terminal-bench#50tau2-bench#64mmmu-pro#731from $0.28 out——
kimi-k3-lowOpen weightsMoonshot AI—unclassified—1.05M—scicode#51mmmu-pro#53artificial-analysis-intelligence-index#600——
Hy3 previewOpen weightsTencent · textOther298.8B262.1K13 Apr 2026tau2-bench#52terminal-bench#91gpqa#1131from $0.60 out✗ too large172.3 GB4bit✗ too large344.1 GB8bit
deepseek-v4-flashOpen weightsDeepSeek—unclassified—1M—livebench-2#54gpqa#56artificial-analysis-intelligence-index#580——
ring-2-6-1tOpen weightsinclusionAI—unclassified—262.1K—tau2-bench#54scicode#98terminal-bench#1240——
Nemotron 3 SuperOpen weightsNVIDIA · text—unclassified—262.1K11 Mar 2026ifbench#55terminal-bench#124scicode#1411from $0.40 out——
Qwen3.5-4BOpen weightsQwenApache-2.04.66B262.1K27 Feb 2026tau2-bench#55mmmu-pro#149ifbench#1700✓ fits3.2 GB4bit✓ fits5.9 GB8bit
QwQ-32BOpen weightsAlibaba Group · textApache-2.032B131.1K6 Mar 2025aider-polyglot#56ifbench#290humanitys-last-exam#3290✓ fits18.9 GB4bit✓ fits37.3 GB8bit
glm-4-7-flash-non-reasoningOpen weightsZ.ai (Zhipu AI)—unclassified—200K—tau2-bench#58ifbench#204terminal-bench#3100——
Qwen3.8 27BOpen weightsQwen · image, text, videoApache-2.027.8B1M5 Aug 2026gpqa#60artificial-analysis-intelligence-index#64mmmu-pro#662from $2.55 out✓ fits16.5 GB4bit✓ fits32.5 GB8bit
Llama 4 MaverickOpen weightsMeta AI · image, text—unclassified—1.05M5 Apr 2025aider-polyglot#61scicode#152mmmu-pro#1721from $0.696 out——
deepseek-v4-pro-0424-non-reasoningOpen weightsDeepSeek—unclassified—1M—tau2-bench#61terminal-bench#72artificial-analysis-intelligence-index#1800——
nvidia-nemotron-3-nano-30b-a3b-reasoningOpen weightsNVIDIA—unclassified—1M—ifbench#62scicode#155terminal-bench#2130——
Gemma 3 27BOpen weightsGoogle · image, text—unclassified—131.1K12 Mar 2025aider-polyglot#67scicode#164mmmu-pro#2241from $0.45 out——
Trinity Large ThinkingOpen weightsArcee AI · text—unclassified—262.1K1 Apr 2026tau2-bench#68scicode#119ifbench#1471from $0.80 out——
motif-3Open weightsMotif Technologies—unclassified—262.1K—artificial-analysis-intelligence-index#68humanitys-last-exam#76scicode#1120——
ling-2-6-1tOpen weightsinclusionAI—unclassified—262.1K—tau2-bench#70terminal-bench#111ifbench#1440——
Hy3Open weightsTencent · textApache-2.0298.8B262.1K2 Jul 2026gpqa#72scicode#81humanitys-last-exam#961from $0.528 out✗ too large172.3 GB4bit✗ too large344.1 GB8bit
MiniMax M2.1Open weightsMiniMax · textOther228.7B204.8K20 Dec 2025ifbench#72tau2-bench#97terminal-bench#1241from $1.2 out✗ too large132.0 GB4bit✗ too large263.5 GB8bit
Inkling SmallOpen weightsThinking Machines · audio, image, text—unclassified—1.05M30 Jul 2026scicode#75gpqa#76mmmu-pro#901from $1.2 out——
mimo-v2-5-pro-non-reasoningOpen weightsXiaomi—unclassified—1M—terminal-bench#75tau2-bench#156artificial-analysis-intelligence-index#2080——
qwen3-5-397b-a17b-non-reasoningOpen weightsAlibaba Group—unclassified—262.1K—terminal-bench#75tau2-bench#109gpqa#1210——
apriel-v1-6-15b-thinkerOpen weightsServiceNow—unclassified—128K—ifbench#76tau2-bench#165terminal-bench#2080——
k2-horizon-375b-a23bOpen weightsMBZUAI Institute of Foundation Models—unclassified—524.3K—artificial-analysis-intelligence-index#80humanitys-last-exam#103gpqa#1040——
minicpm-v4-6-1-3bOpen weightsOpenBMB—unclassified—262K—tau2-bench#80mmmu-pro#243terminal-bench#3780——
qwen3-5-4b-non-reasoningOpen weightsAlibaba Group—unclassified—262.1K—tau2-bench#80mmmu-pro#174terminal-bench#2260——
hyperclova-x-seed-think-32bOpen weightsNaver—unclassified—128K—tau2-bench#82terminal-bench#221ifbench#3010——
deepseek-v3-2-specialeOpen weightsDeepSeek—unclassified—128K—terminal-bench#83gpqa#108ifbench#1130——
nex-n2-proOpen weightsNex AGI—unclassified—262K—gpqa#83humanitys-last-exam#91artificial-analysis-intelligence-index#960——
qwen3-5-27b-non-reasoningOpen weightsAlibaba Group—unclassified—262.1K—tau2-bench#83terminal-bench#105mmmu-pro#1240——
solar-open2-250bOpen weightsUpstage—unclassified—1.05M—scicode#84humanitys-last-exam#122gpqa#1270——
Muse Glimmer 30BOpen weightsMeta AI · image, text—unclassified—131.1K9 Aug 2026mmmu-pro#86scicode#99humanitys-last-exam#1613from $0.75 out——
Qwen3.5-9BOpen weightsQwen · image, text, videoApache-2.09.65B262.1K27 Feb 2026tau2-bench#86ifbench#92mmmu-pro#1282from $0.15 out✓ fits6.0 GB4bit✓ fits11.6 GB8bit
InklingOpen weightsThinking Machines · audio, image, text—unclassified—1.05M17 Jul 2026scicode#88mmmu-pro#97humanitys-last-exam#1042from $4.05 out——
qwen3-8-27b-mediumOpen weightsAlibaba Group—unclassified—256K—mmmu-pro#89artificial-analysis-intelligence-index#101scicode#1260——
qwen3-5-35b-a3b-non-reasoningOpen weightsAlibaba Group—unclassified—262.1K—tau2-bench#91mmmu-pro#131gpqa#1850——
LongCat 2.0Open weightsMeituan · text—unclassified—1.05M20 Jul 2026humanitys-last-exam#92scicode#140artificial-analysis-intelligence-index#1921from $1.2 out——
agnes-2-5-pro-alphaOpen weightsSapiens AI—unclassified—1M—humanitys-last-exam#94gpqa#100artificial-analysis-intelligence-index#1020——

Best results = rank inside each benchmark's primary comparability group; several dimensions are shown side by side and never combined. Fit columns are estimates (64 GB device at 4-bit, 128 GB at 8-bit, 8K context) — method on /methodology. unclassified = the raw licence label is not yet mapped in the ontology.

MethodUniverse = canonical models whose weights can be downloaded (open-weights, open-source, restricted-weights). hardware_fit values are ESTIMATES (see /methodology); best_results are ranks inside each benchmark's primary comparability group — no composite score. /methodology