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.

889 downloadable models

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 (311) · MIT License (125) · Other (unclassified licence) (83) · Creative Commons Attribution-NonCommercial 4.0 (24) · Apple Sample Code / ML Research License (15) · Llama 3.2 Community License (12)

Category · this pageCommercial useRedistributionDerivativesHosting
Open source n=0
Open weights n=5029120300203002029120
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

889 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
Ming-flash-omni-2.0Open weightsinclusionAIMIT104.2B10 Feb 20260✓ fits60.4 GB4bit✓ fits120.4 GB8bit
Ming-omni-tts-0.5BOpen weightsinclusionAI · audio, textApache-2.01.62B11 Feb 20260✓ fits1.4 GB4bit✓ fits2.4 GB8bit
MiniMax-01Open weightsMiniMax · textunclassified456.1B1M12 Jan 20252from $1.1 out✗ too large262.8 GB4bit✗ too large525.0 GB8bit
MiniMax-H3Open weightsMiniMax · image, text, videoOther33.1B28 Jul 20260✓ fits19.5 GB4bit✓ fits38.6 GB8bit
MiniMax-M1-40k-hfOpen weightsMiniMax · textApache-2.0456.1B1 Jul 20250✗ too large262.8 GB4bit✗ too large525.0 GB8bit
MiniMax-M1-80k-hfOpen weightsMiniMax · textApache-2.0456.1B1 Jul 20250✗ too large262.8 GB4bit✗ too large525.0 GB8bit
MiniMax-Music3Open weightsMiniMax · audio, textunclassified2.43B7 Aug 20260✓ fits1.9 GB4bit✓ fits3.3 GB8bit
MiniMax-Text-01-hfOpen weightsMiniMax · textOther456.1B3 Jun 20250✗ too large262.8 GB4bit✗ too large525.0 GB8bit
MiniMax-VL-01Open weightsMiniMax · image, textunclassified456.4B12 Jan 20250✗ too large262.9 GB4bit✗ too large525.4 GB8bit
Ministral 3 14B 2512Open weightsMistral AI · image, textApache-2.013.9B262.1K31 Oct 20252from $0.20 out✓ fits8.5 GB4bit✓ fits16.5 GB8bit
Ministral 3 3B 2512Open weightsMistral AI · image, textApache-2.03.85B131.1K31 Oct 20252from $0.10 out✓ fits2.7 GB4bit✓ fits4.9 GB8bit
Ministral 3 8B 2512Open weightsMistral AI · image, textunclassified262.1K2 Dec 20252from $0.075 out
Ministral-3-14B-Base-2512Open weightsMistral AIunclassified0
Ministral-3-14B-Reasoning-2512Open weightsMistral AIApache-2.013.9B31 Oct 20250✓ fits8.5 GB4bit✓ fits16.5 GB8bit
Ministral-3-3B-Base-2512Open weightsMistral AIunclassified0
Ministral-3-3B-Reasoning-2512Open weightsMistral AIApache-2.04.25B31 Oct 20250✓ fits3.0 GB4bit✓ fits5.4 GB8bit
Mistral NemoOpen weightsMistral AI · textApache-2.012.3B131.1K17 Jul 20242from $0.03 out✓ fits7.5 GB4bit✓ fits14.6 GB8bit
Mistral Small 3.1 24BOpen weightsMistral AI · image, textApache-2.024B128K11 Mar 20252from $0.555 out✓ fits14.3 GB4bit✓ fits28.1 GB8bit
Mistral Small 3.2 24BOpen weightsMistral AI · image, textApache-2.024B256K19 Jun 20252from $0.20 out✓ fits14.3 GB4bit✓ fits28.1 GB8bit
Mistral-7B-Instruct-v0.1Open weightsMistral AI · textApache-2.07.24B27 Sept 20230✓ fits4.7 GB4bit✓ fits8.8 GB8bit
Mistral-7B-Instruct-v0.2Open weightsMistral AI · textApache-2.07.24B11 Dec 20230✓ fits4.7 GB4bit✓ fits8.8 GB8bit
Mistral-7B-Instruct-v0.3Open weightsMistral AIApache-2.07.25B22 May 20240✓ fits4.7 GB4bit✓ fits8.8 GB8bit
Mistral-7B-v0.1Open weightsMistral AI · textApache-2.07.24B20 Sept 20230✓ fits4.7 GB4bit✓ fits8.8 GB8bit
Mistral-7B-v0.3Open weightsMistral AIApache-2.07.25B22 May 20240✓ fits4.7 GB4bit✓ fits8.8 GB8bit
Mistral-Medium-3.5-128BOpen weightsMistral AIOther127.7B31 Mar 20260✗ too large73.9 GB4bit✗ too large147.4 GB8bit
Mistral-Nemo-Base-2407Open weightsMistral AIApache-2.012.3B18 Jul 20240✓ fits7.5 GB4bit✓ fits14.6 GB8bit
Mistral-Small-24B-Base-2501Open weightsMistral AIunclassified0
Mistral-Small-3.1-24B-Base-2503Open weightsMistral AIunclassified0
Mixtral 8x22B InstructOpen weightsMistral AI · document, textunclassified65.5K17 Apr 20242from $6 out
Mixtral-8x7B-Instruct-v0.1Open weightsMistral AIApache-2.046.7B10 Dec 20230✓ fits27.4 GB4bit✓ fits54.2 GB8bit
Mixtral-8x7B-v0.1Open weightsMistral AIunclassified0
Molmo2-4BOpen weightsAllen Institute for AI · image, textApache-2.04.85B14 Dec 20250✓ fits3.3 GB4bit✓ fits6.1 GB8bit
MoonViT-SO-400MOpen weightsMoonshot AI · embedding, imageMIT416.9M10 Apr 20250✓ fits0.7 GB4bit✓ fits1.0 GB8bit
Moonlight-16B-A3BOpen weightsMoonshot AI · textMIT16B3B active22 Feb 20250✓ fits9.7 GB4bit✓ fits18.9 GB8bit
Moonlight-16B-A3B-InstructOpen weightsMoonshot AI · textMIT16B3B active22 Feb 20250✓ fits9.7 GB4bit✓ fits18.9 GB8bit
MythoMax 13BOpen weightsgryphe · textunclassified8.19K2 Jul 20231from $0.06 out
Nemotron 3 Nano Omni (free)Open weightsNVIDIAunclassified2
Nemotron 3.5 Content SafetyOpen weightsNVIDIA · image, textunclassified131.1K4 Jun 20262from $0.20 out
Nex-N2.5-Mini (free)Open weightsNex AGIunclassified1
Nex-N2.5-Pro (free)Open weightsNex AGIunclassified1
NextStep-1.1Open weightsStepFun · image, textApache-2.014.9B23 Dec 20250✓ fits9.1 GB4bit✓ fits17.7 GB8bit
North Small TranslateOpen weightsCohere · textCC-BY-NC-4.0218B25B active16K10 Sept 20260✗ too large125.8 GB4bit✗ too large251.2 GB8bit
North-Micro-Vision-InstructOpen weightsCohere · image, textApache-2.02.48B10 Aug 20260✓ fits1.9 GB4bit✓ fits3.4 GB8bit
Nous-Hermes-2-Mixtral-8x7B-DPOOpen weightsNous Research · textApache-2.046.7B11 Jan 20240✓ fits27.4 GB4bit✓ fits54.2 GB8bit
Nous-Hermes-2-SOLAR-10.7BOpen weightsNous Research · textApache-2.010.7B1 Jan 20240✓ fits6.7 GB4bit✓ fits12.8 GB8bit
Nous-Hermes-2-Yi-34BOpen weightsNous Research · textApache-2.034.4B23 Dec 20230✓ fits20.3 GB4bit✓ fits40.0 GB8bit
Nous-Hermes-llama-2-7bOpen weightsNous Research · textMIT6.74B25 Jul 20230✓ fits4.4 GB4bit✓ fits8.3 GB8bit
NousResearch/Llama-2-7b-chat-hfOpen weightsNous Research · textunclassified6.74B18 Jul 20230✓ fits4.4 GB4bit✓ fits8.3 GB8bit
NousResearch/Llama-2-7b-hfOpen weightsNous Research · textunclassified6.74B18 Jul 20230✓ fits4.4 GB4bit✓ fits8.3 GB8bit
OASIS-code-1.3BOpen weightsKwaipilot · embedding, textMIT1.28B30 Oct 20240✓ fits1.2 GB4bit✓ fits2.0 GB8bit

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