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 | 40010 | 40010 | 40010 | 40010 |
| 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 | |
|---|---|---|---|---|---|---|---|---|
| SmolVLM-256M-InstructOpen weightsHugging Face · image, text | Apache-2.0 | 256.5M | — | 17 Jan 2025 | — | 0 | ✓ fits0.7 GB4bit✓ fits0.8 GB8bit | |
| SmolVLM-500M-InstructOpen weightsHugging Face · image, text | Apache-2.0 | 507.5M | — | 20 Jan 2025 | — | 0 | ✓ fits0.8 GB4bit✓ fits1.1 GB8bit | |
| SmolVLM-InstructOpen weightsHugging Face · image, text | Apache-2.0 | 2.25B | — | 18 Nov 2024 | — | 0 | ✓ fits1.8 GB4bit✓ fits3.1 GB8bit | |
| SmolVLM2-2.2B-InstructOpen weightsHugging Face · image, text | Apache-2.0 | 2.25B | — | 8 Feb 2025 | — | 0 | ✓ fits1.8 GB4bit✓ fits3.1 GB8bit | |
| SmolVLM2-256M-Video-InstructOpen weightsHugging Face · image, text | Apache-2.0 | 256.5M | — | 11 Feb 2025 | — | 0 | ✓ fits0.7 GB4bit✓ fits0.8 GB8bit | |
| SmolVLM2-500M-Video-InstructOpen weightsHugging Face · image, text | Apache-2.0 | 507.5M | — | 11 Feb 2025 | — | 0 | ✓ fits0.8 GB4bit✓ fits1.1 GB8bit | |
| Stable-DiffCoder-8B-BaseOpen weightsByteDance · text | MIT | 8.25B | — | 15 Jan 2026 | — | 0 | ✓ fits5.3 GB4bit✓ fits10.0 GB8bit | |
| Stable-DiffCoder-8B-InstructOpen weightsByteDance · text | MIT | 8.25B | — | 15 Jan 2026 | — | 0 | ✓ fits5.3 GB4bit✓ fits10.0 GB8bit | |
| Step-3.5-Flash-BaseOpen weightsStepFun · text | Apache-2.0 | 197.8B | — | 2 Mar 2026 | — | 0 | ✗ too large114.2 GB4bit✗ too large227.9 GB8bit | |
| Step-3.5-Flash-REAP-121B-A11BOpen weightsCerebras Systems · text | Other | 121B11B active | — | 25 Feb 2026 | — | 0 | ✗ too large70.0 GB4bit✗ too large139.6 GB8bit | |
| Step-3.5-Flash-REAP-149B-A11BOpen weightsCerebras Systems · text | Other | 149.4B11B active | — | 25 Feb 2026 | — | 0 | ✗ too large86.4 GB4bit✗ too large172.3 GB8bit | |
| Step-Audio-2-miniOpen weightsStepFun | Apache-2.0 | 8.32B | — | 28 Aug 2025 | — | 0 | ✓ fits5.3 GB4bit✓ fits10.1 GB8bit | |
| Step-Audio-R1.1Open weightsStepFun · audio, text | Apache-2.0 | 33.5B | — | 14 Jan 2026 | — | 0 | ✓ fits19.8 GB4bit✓ fits39.0 GB8bit | |
| Step1X-Edit-v1p2Open weightsStepFun · image | Apache-2.0 | 12.4B | — | 26 Nov 2025 | — | 0 | ✓ fits7.7 GB4bit✓ fits14.8 GB8bit | |
| StepFun-Formalizer-7BOpen weightsStepFun · text | Apache-2.0 | 7.62B | — | 19 Aug 2025 | — | 0 | ✓ fits4.9 GB4bit✓ fits9.3 GB8bit | |
| StripedHyena-Hessian-7BOpen weightsTogether AI · text | Apache-2.0 | 7.65B | — | 21 Nov 2023 | — | 0 | ✓ fits4.9 GB4bit✓ fits9.3 GB8bit | |
| StripedHyena-Nous-7BOpen weightsTogether AI · text | Apache-2.0 | 7.65B | — | 4 Dec 2023 | — | 0 | ✓ fits4.9 GB4bit✓ fits9.3 GB8bit | |
| StyleTTS2-LJSpeechOpen weightsyl4579 | —unclassified | — | — | — | — | 0 | —— | |
| SynLogic-32BOpen weightsMiniMax · text | MIT | 32.8B | — | 30 May 2025 | — | 0 | ✓ fits19.3 GB4bit✓ fits38.2 GB8bit | |
| SynLogic-7BOpen weightsMiniMax · text | MIT | 7.62B | — | 3 Jun 2025 | — | 0 | ✓ fits4.9 GB4bit✓ fits9.3 GB8bit | |
| SynLogic-Mix-3-32BOpen weightsMiniMax · text | MIT | 32.8B | — | 30 May 2025 | — | 0 | ✓ fits19.3 GB4bit✓ fits38.2 GB8bit | |
| TRELLIS-image-largeOpen weightsMicrosoft · 3d, image | MIT | — | — | 2 Dec 2024 | — | 0 | —— | |
| TRELLIS.2-4BOpen weightsMicrosoft · 3d, image | MIT | 4B | — | 1 Dec 2025 | — | 0 | ✓ fits2.8 GB4bit✓ fits5.1 GB8bit | |
| Thomson-1.0-SmallOpen weightsthomsonreuters | —unclassified | — | — | — | — | 0 | —— | |
| TinySolar-111m-4kOpen weightsUpstage | Apache-2.0 | 111.2M | — | 27 May 2026 | — | 0 | ✓ fits0.6 GB4bit✓ fits0.6 GB8bit | |
| TinySolar-187m-4kOpen weightsUpstage | Apache-2.0 | 187.2M | — | 27 May 2026 | — | 0 | ✓ fits0.6 GB4bit✓ fits0.7 GB8bit | |
| TinySolar-248m-4kOpen weightsUpstage · text | Apache-2.0 | 248M | — | 7 Feb 2024 | — | 0 | ✓ fits0.6 GB4bit✓ fits0.8 GB8bit | |
| TinySolar-248m-4k-code-instructOpen weightsUpstage · text | Apache-2.0 | 248M | — | 15 Apr 2024 | — | 0 | ✓ fits0.6 GB4bit✓ fits0.8 GB8bit | |
| TinySolar-248m-4k-pyOpen weightsUpstage · text | Apache-2.0 | 248M | — | 7 Feb 2024 | — | 0 | ✓ fits0.6 GB4bit✓ fits0.8 GB8bit | |
| TinySolar-248m-4k-py-instructOpen weightsUpstage · text | Apache-2.0 | 248M | — | 11 Apr 2024 | — | 0 | ✓ fits0.6 GB4bit✓ fits0.8 GB8bit | |
| Titan-text-embeddings-v2Open weightsAmazon Web Services · embedding, text | Other | — | — | 30 Apr 2024 | — | 0 | —— | |
| TripoSROpen weightsStability AI · 3d, image | MIT | — | — | 29 Feb 2024 | — | 0 | —— | |
| UI-Mate-9BOpen weightsTencent | —unclassified | — | — | — | — | 0 | —— | |
| UI-TARS 7BOpen weightsByteDance · image, text | Apache-2.0 | 8.29B | 128K | 16 Apr 2025 | — | 1from $0.20 out | ✓ fits5.3 GB4bit✓ fits10.0 GB8bit | |
| UI-TARS-2B-SFTOpen weightsByteDance · image, text | Apache-2.0 | 2.44B | — | 20 Jan 2025 | — | 0 | ✓ fits1.9 GB4bit✓ fits3.3 GB8bit | |
| UI-TARS-72B-DPOOpen weightsByteDance · image, text | Apache-2.0 | 73.4B | — | 22 Jan 2025 | — | 0 | ✓ fits42.7 GB4bit✓ fits84.9 GB8bit | |
| UI-TARS-7B-DPOOpen weightsByteDance · image, text | Apache-2.0 | 8.29B | — | 22 Jan 2025 | — | 0 | ✓ fits5.3 GB4bit✓ fits10.0 GB8bit | |
| UI-TARS-7B-SFTOpen weightsByteDance · image, text | Apache-2.0 | 8.29B | — | 20 Jan 2025 | — | 0 | ✓ fits5.3 GB4bit✓ fits10.0 GB8bit | |
| UI-Venus-1.5-2BOpen weightsinclusionAI · image, text | Apache-2.0 | 2.44B | — | 9 Feb 2026 | — | 0 | ✓ fits1.9 GB4bit✓ fits3.3 GB8bit | |
| UncensoredOpen weightsVenice · text | —unclassified | — | 128K | 9 Jul 2025 | — | 1from $0.90 out | —— | |
| Unlimited-OCROpen weightsBaidu · image, text | MIT | 3.34B | — | 19 Jun 2026 | — | 0 | ✓ fits2.4 GB4bit✓ fits4.3 GB8bit | |
| UnslopNemo 12BOpen weightsTheDrummer · text | —unclassified | — | 1.02M | 8 Nov 2024 | — | 1from $0.40 out | —— | |
| VTP-Large-f16d64Open weightsMiniMax · embedding, image | Other | 731.6M | — | 16 Dec 2025 | — | 0 | ✓ fits0.9 GB4bit✓ fits1.3 GB8bit | |
| VTP-Small-f16d64Open weightsMiniMax · embedding, image | Other | 167.2M | — | 16 Dec 2025 | — | 0 | ✓ fits0.6 GB4bit✓ fits0.7 GB8bit | |
| VibeVoice-ASROpen weightsMicrosoft · audio, text | MIT | 8.67B | — | 21 Jan 2026 | — | 0 | ✓ fits5.5 GB4bit✓ fits10.5 GB8bit | |
| Voxtral Mini Transcribe RealtimeOpen weightsMistral AI | Apache-2.0 | — | — | — | — | 0 | —— | |
| Voxtral SmallOpen weightsMistral AI | Apache-2.0 | — | 32K | — | — | 1from $0.40 out | —— | |
| Voxtral Small 24B 2507Open weightsMistral AI · audio, document, text | Apache-2.0 | 24.3B | 32.8K | 1 Jul 2025 | — | 2from $0.30 out | ✓ fits14.4 GB4bit✓ fits28.4 GB8bit | |
| Voxtral-Mini-3B-2507Open weightsMistral AI | Apache-2.0 | 4.68B | — | 1 Jul 2025 | — | 0 | ✓ fits3.2 GB4bit✓ fits5.9 GB8bit | |
| Voxtral-Mini-4B-Realtime-2602Open weightsMistral AI · audio, text | Apache-2.0 | 4.43B | — | 21 Jan 2026 | — | 0 | ✓ fits3.0 GB4bit✓ fits5.6 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