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.
- Open source1 0%
- Open weights889 87%
- Restricted weights129 13%
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 page | Commercial use | Redistribution | Derivatives | Hosting |
|---|---|---|---|---|
| Open source n=0 | — | — | — | — |
| Open weights n=50 | 34016 | 34016 | 34016 | 34016 |
| 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
| Model | Licence · permissions | Params | Context | Release | Best results | Providers | Fit 64 GB @4bit · 128 GB @8bit Estimated | |
|---|---|---|---|---|---|---|---|---|
| Fara1.5-27BOpen weightsMicrosoft | —unclassified | — | — | — | — | 0 | —— | |
| Fara1.5-4BOpen weightsMicrosoft | —unclassified | — | — | — | — | 0 | —— | |
| Fara1.5-9BOpen weightsMicrosoft | —unclassified | — | — | — | — | 0 | —— | |
| Florence-2-baseOpen weightsMicrosoft · image, text | MIT | 231.6M | — | 15 Jun 2024 | — | 0 | ✓ fits0.6 GB4bit✓ fits0.8 GB8bit | |
| Florence-2-largeOpen weightsMicrosoft · image, text | MIT | 776.7M | — | 15 Jun 2024 | — | 0 | ✓ fits0.9 GB4bit✓ fits1.4 GB8bit | |
| GDN-primed-HQwen3-8B-InstructOpen weightsAmazon Web Services · text | Apache-2.0 | 8.5B | — | 31 Mar 2026 | — | 0 | ✓ fits5.4 GB4bit✓ fits10.3 GB8bit | |
| GELab-Zero-4B-previewOpen weightsStepFun · image, text | Apache-2.0 | 4.44B | — | 28 Nov 2025 | — | 0 | ✓ fits3.0 GB4bit✓ fits5.6 GB8bit | |
| GKA-primed-HQwen3-32B-ReasonerOpen weightsAmazon Web Services · text | Apache-2.0 | 34.1B | — | 31 Mar 2026 | — | 0 | ✓ fits20.1 GB4bit✓ fits39.8 GB8bit | |
| GKA-primed-HQwen3-8B-ReasonerOpen weightsAmazon Web Services · text | Apache-2.0 | 8.5B | — | 31 Mar 2026 | — | 0 | ✓ fits5.4 GB4bit✓ fits10.3 GB8bit | |
| GLM-4-9B-0414Open weightsZ.ai (Zhipu AI) | —unclassified | — | — | — | — | 0 | —— | |
| GLM-4.1V-9B-ThinkingOpen weightsZ.ai (Zhipu AI) · image, text | MIT | 10.3B | — | 28 Jun 2025 | — | 0 | ✓ fits6.4 GB4bit✓ fits12.3 GB8bit | |
| GLM-4.6V-FlashOpen weightsZ.ai (Zhipu AI) · image, text | MIT | 10.3B | — | 7 Dec 2025 | — | 0 | ✓ fits6.4 GB4bit✓ fits12.3 GB8bit | |
| GLM-ASR-Nano-2512Open weightsZ.ai (Zhipu AI) · audio, text | MIT | 2.26B | — | 9 Dec 2025 | — | 0 | ✓ fits1.8 GB4bit✓ fits3.1 GB8bit | |
| GLM-OCROpen weightsZ.ai (Zhipu AI) · image, text | MIT | 1.33B | — | 30 Jan 2026 | — | 0 | ✓ fits1.3 GB4bit✓ fits2.0 GB8bit | |
| GOT-OCR-2.0-hfOpen weightsStepFun · image, text | Apache-2.0 | 560.5M | — | 22 Nov 2024 | — | 0 | ✓ fits0.8 GB4bit✓ fits1.1 GB8bit | |
| GOT-OCR2_0Open weightsStepFun · image, text | Apache-2.0 | 716M | — | 12 Sept 2024 | — | 0 | ✓ fits0.9 GB4bit✓ fits1.3 GB8bit | |
| GPT-JT-6B-v1Open weightsTogether AI | Apache-2.0 | 6B | — | 24 Nov 2022 | — | 0 | ✓ fits4.0 GB4bit✓ fits7.4 GB8bit | |
| GPT-NeoXT-Chat-Base-20BOpen weightsTogether AI · text | Apache-2.0 | 20B | — | 3 Mar 2023 | — | 0 | ✓ fits12.0 GB4bit✓ fits23.5 GB8bit | |
| Gemma 2 27BOpen weightsGoogle · text | —unclassified | — | 8.19K | 13 Jul 2024 | — | 2from $0.65 out | —— | |
| Grug-12BOpen weightskai-os | —unclassified | — | — | — | — | 0 | —— | |
| Hermes 3 405B InstructOpen weightsNous Research · text | —unclassified | — | 131.1K | 16 Aug 2024 | — | 1from $1 out | —— | |
| Hermes-2-Pro-Llama-3-8BOpen weightsNous Research | —unclassified | — | — | — | — | 0 | —— | |
| Hermes-2-Theta-Llama-3-8BOpen weightsNous Research · text | Apache-2.0 | 8.03B | — | 5 May 2024 | — | 0 | ✓ fits5.1 GB4bit✓ fits9.7 GB8bit | |
| HiPO-1.7BOpen weightsKwaipilot · text | Apache-2.0 | 2.03B | — | 31 Oct 2025 | — | 0 | ✓ fits1.7 GB4bit✓ fits2.8 GB8bit | |
| HiPO-8BOpen weightsKwaipilot · text | Apache-2.0 | 8.19B | — | 26 Sept 2025 | — | 0 | ✓ fits5.2 GB4bit✓ fits9.9 GB8bit | |
| Hunyuan-7B-InstructOpen weightsTencent · text | —unclassified | 7.5B | — | 30 Jul 2025 | — | 0 | ✓ fits4.8 GB4bit✓ fits9.1 GB8bit | |
| Hunyuan-7B-PretrainOpen weightsTencent | —unclassified | — | — | — | — | 0 | —— | |
| HunyuanImage-3.0-InstructOpen weightsTencent · image | Other | 83B | — | 25 Sept 2025 | — | 0 | ✓ fits48.2 GB4bit✓ fits96.0 GB8bit | |
| Hy-MT2-1.8BOpen weightsTencent · text | Apache-2.0 | 2.04B | 8.19K | 11 May 2026 | — | 1from $0.177 out | ✓ fits1.7 GB4bit✓ fits2.9 GB8bit | |
| Hy-MT2-30B-A3BOpen weightsTencent · text | Apache-2.0 | 30.1B3B active | 8.19K | 11 May 2026 | — | 1from $0.295 out | ✓ fits17.8 GB4bit✓ fits35.1 GB8bit | |
| Hy-MT2-7BOpen weightsTencent · text | Apache-2.0 | 8.03B | 8.19K | 11 May 2026 | — | 1from $0.295 out | ✓ fits5.1 GB4bit✓ fits9.7 GB8bit | |
| Hy4 previewOpen weightsTencent · text | —unclassified | — | 1.05M | 28 Aug 2026 | — | 1from $2.5 out | —— | |
| Intern-S1Open weightsInternLM (Shanghai AI Laboratory) · image, text | Apache-2.0 | 240.7B | — | 24 Jul 2025 | — | 0 | ✗ too large138.9 GB4bit✗ too large277.3 GB8bit | |
| Intern-S1-ProOpen weightsInternLM (Shanghai AI Laboratory) · image, text | Apache-2.0 | — | — | 2 Feb 2026 | — | 0 | —— | |
| Intern-S1-miniOpen weightsInternLM (Shanghai AI Laboratory) · image, text | Apache-2.0 | 8.54B | — | 18 Aug 2025 | — | 0 | ✓ fits5.4 GB4bit✓ fits10.3 GB8bit | |
| InternVL3-1B-InstructOpen weightsOpenGVLab | —unclassified | — | — | — | — | 0 | —— | |
| Jamba-tiny-devOpen weightsAI21 Labs | Apache-2.0 | 318.7M | — | 3 Sept 2024 | — | 0 | ✓ fits0.7 GB4bit✓ fits0.9 GB8bit | |
| Jamba-tiny-randomOpen weightsAI21 Labs · text | Apache-2.0 | 127.7M | — | 17 Apr 2024 | — | 0 | ✓ fits0.6 GB4bit✓ fits0.7 GB8bit | |
| Jamba-tiny-reward-devOpen weightsAI21 Labs | Apache-2.0 | 318.7M | — | 5 Dec 2024 | — | 0 | ✓ fits0.7 GB4bit✓ fits0.9 GB8bit | |
| Jamba-v0.1Open weightsAI21 Labs · text | Apache-2.0 | 51.6B | — | 28 Mar 2024 | — | 0 | ✓ fits30.2 GB4bit✓ fits59.8 GB8bit | |
| K-EXAONE-2.0-750B-A37BOpen weightsLG AI Research · text | Apache-2.0 | 749.4B37B active | — | 29 Jul 2026 | — | 0 | ✗ too large431.4 GB4bit✗ too large862.3 GB8bit | |
| K-EXAONE-236B-A23BOpen weightsLG AI Research · text | Other | 237.1B23B active | — | 26 Dec 2025 | — | 0 | ✗ too large136.8 GB4bit✗ too large273.2 GB8bit | |
| KAT-Coder-V2.5-DevOpen weightsKwaipilot · text | Apache-2.0 | 34.7B | — | 23 Jul 2026 | — | 0 | ✓ fits20.4 GB4bit✓ fits40.4 GB8bit | |
| KAT-DevOpen weightsKwaipilot · text | Apache-2.0 | 32.8B | — | 15 Sept 2025 | — | 0 | ✓ fits19.3 GB4bit✓ fits38.2 GB8bit | |
| KAT-Dev-72B-ExpOpen weightsKwaipilot · text | Apache-2.0 | 72.7B | — | 10 Oct 2025 | — | 0 | ✓ fits42.3 GB4bit✓ fits84.1 GB8bit | |
| KAT-V1-40BOpen weightsKwaipilot · text | Other | 40.6B | — | 20 Jul 2025 | — | 0 | ✓ fits23.8 GB4bit✓ fits47.1 GB8bit | |
| KaLM-Embedding-Gemma3-12B-2511Open weightsTencent · embedding, text | Other | 11.8B | — | 4 Nov 2025 | — | 0 | ✓ fits7.3 GB4bit✓ fits14.0 GB8bit | |
| Kimi-Audio-7BOpen weightsMoonshot AI · audio, text | MIT | 9.77B | — | 25 Apr 2025 | — | 0 | ✓ fits6.1 GB4bit✓ fits11.7 GB8bit | |
| Kimi-Audio-7B-InstructOpen weightsMoonshot AI · audio, text | MIT | 9.77B | — | 25 Apr 2025 | — | 0 | ✓ fits6.1 GB4bit✓ fits11.7 GB8bit | |
| Kimi-Dev-72BOpen weightsMoonshot AI · text | MIT | 72.7B | — | 16 Jun 2025 | — | 0 | ✓ fits42.3 GB4bit✓ fits84.1 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