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.

950 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 (275) · Other (unclassified licence) (115) · MIT License (92) · 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=4228014280142801428014
restricted-weights n=8080800800080

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

Downloadable models

950 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
aya-101Open weightsCohereApache-2.012.9B8 Feb 20240✓ fits7.9 GB4bit✓ fits15.4 GB8bit
aya-23-8Brestricted-weightsCohereCC-BY-NC-4.08.03B19 May 20240✓ fits5.1 GB4bit✓ fits9.7 GB8bit
aya-expanse-32brestricted-weightsCohereCC-BY-NC-4.032.3B23 Oct 20240✓ fits19.1 GB4bit✓ fits37.6 GB8bit
aya-expanse-8brestricted-weightsCohereCC-BY-NC-4.08.03B23 Oct 20240✓ fits5.1 GB4bit✓ fits9.7 GB8bit
aya-vision-8brestricted-weightsCohereCC-BY-NC-4.08.63B2 Mar 20250✓ fits5.5 GB4bit✓ fits10.4 GB8bit
battlestar-gpt2-small-x49Open weightsStanford CRFMunclassified2 Mar 20220
beren-gpt2-medium-x49Open weightsStanford CRFMunclassified2 Mar 20220
bert-base-uncasedOpen weightsgoogle-bertunclassified0
bigvgan_v2_22khz_80band_256xOpen weightsNVIDIAMIT15 Jul 20240
bigvgan_v2_44khz_128band_512xOpen weightsNVIDIAMIT15 Jul 20240
blip-image-captioning-baseOpen weightsSalesforceBSD-3-Clause12 Dec 20220
blip-image-captioning-largeOpen weightsSalesforceBSD-3-Clause469.7M13 Dec 20220✓ fits0.8 GB4bit✓ fits1.0 GB8bit
blip-itm-base-cocoOpen weightsSalesforceBSD-3-Clause12 Dec 20220
blip-vqa-baseOpen weightsSalesforceBSD-3-Clause384.7M12 Dec 20220✓ fits0.7 GB4bit✓ fits0.9 GB8bit
blip2-flan-t5-xlOpen weightsSalesforceMIT3.94B6 Feb 20230✓ fits2.8 GB4bit✓ fits5.0 GB8bit
blip2-opt-2.7bOpen weightsSalesforceMIT3.74B6 Feb 20230✓ fits2.6 GB4bit✓ fits4.8 GB8bit
bortOpen weightsAmazon Web Servicesunclassified2 Mar 20220
browsesafeOpen weightsPerplexity AIMIT30.5B14 Nov 20250✓ fits18.1 GB4bit✓ fits35.6 GB8bit
c4ai-command-a-03-2025restricted-weightsCohereCC-BY-NC-4.0111.1B11 Mar 20250✗ too large64.4 GB4bit✗ too large128.2 GB8bit
c4ai-command-r-08-2024restricted-weightsCohereCC-BY-NC-4.032.3B19 Aug 20240✓ fits19.1 GB4bit✓ fits37.6 GB8bit
c4ai-command-r-v01restricted-weightsCohereCC-BY-NC-4.035B11 Mar 20240✓ fits20.6 GB4bit✓ fits40.7 GB8bit
c4ai-command-r7b-12-2024restricted-weightsCohereCC-BY-NC-4.08.03B11 Dec 20240✓ fits5.1 GB4bit✓ fits9.7 GB8bit
caprica-gpt2-small-x81Open weightsStanford CRFMunclassified2 Mar 20220
celebrimbor-gpt2-medium-x81Open weightsStanford CRFMunclassified2 Mar 20220
cerebras/DeepSeek-V3.2-REAP-345B-A37BOpen weightsCerebras SystemsMIT344.9B37B active9 Dec 20250✗ too large198.8 GB4bit✗ too large397.1 GB8bit
cerebras/DeepSeek-V3.2-REAP-508B-A37BOpen weightsCerebras SystemsMIT508.4B37B active9 Dec 20250✗ too large292.8 GB4bit✗ too large585.2 GB8bit
cerebras/GLM-4.5-Air-REAP-82B-A12BOpen weightsCerebras SystemsMIT81.9B12B active20 Oct 20250✓ fits47.6 GB4bit✓ fits94.7 GB8bit
cerebras/GLM-4.6-REAP-218B-A32BOpen weightsCerebras SystemsMIT218.4B32B active23 Oct 20250✗ too large126.1 GB4bit✗ too large251.6 GB8bit
cerebras/GLM-4.6-REAP-268B-A32BOpen weightsCerebras SystemsMIT268.8B32B active24 Oct 20250✗ too large155.1 GB4bit✗ too large309.6 GB8bit
cerebras/GLM-4.7-Flash-REAP-23B-A3BOpen weightsCerebras SystemsMIT23B3B active23 Jan 20260✓ fits13.7 GB4bit✓ fits26.9 GB8bit
cerebras/GLM-4.7-REAP-218B-A32BOpen weightsCerebras SystemsMIT218.4B32B active10 Jan 20260✗ too large126.1 GB4bit✗ too large251.6 GB8bit
cerebras/GLM-4.7-REAP-268B-A32BOpen weightsCerebras SystemsMIT268.8B32B active10 Jan 20260✗ too large155.1 GB4bit✗ too large309.6 GB8bit
cerebras/Kimi-Linear-REAP-35B-A3B-InstructOpen weightsCerebras SystemsMIT35.1B3B active6 Nov 20250✓ fits20.7 GB4bit✓ fits40.9 GB8bit
cerebras/MiniMax-M2-REAP-139B-A10BOpen weightsCerebras SystemsOther139.2B10B active17 Nov 20250✗ too large80.5 GB4bit✗ too large160.5 GB8bit
cerebras/MiniMax-M2-REAP-162B-A10BOpen weightsCerebras SystemsOther162B10B active15 Nov 20250✗ too large93.6 GB4bit✗ too large186.8 GB8bit
cerebras/MiniMax-M2-REAP-172B-A10BOpen weightsCerebras SystemsOther172.5B10B active15 Nov 20250✗ too large99.7 GB4bit✗ too large198.9 GB8bit
cerebras/MiniMax-M2.1-REAP-139B-A10BOpen weightsCerebras SystemsOther139.2B10B active27 Jan 20260✗ too large80.5 GB4bit✗ too large160.5 GB8bit
cerebras/MiniMax-M2.5-REAP-139B-A10BOpen weightsCerebras SystemsOther139.2B10B active18 Feb 20260✗ too large80.5 GB4bit✗ too large160.5 GB8bit
cerebras/MiniMax-M2.5-REAP-172B-A10BOpen weightsCerebras SystemsOther172.5B10B active18 Feb 20260✗ too large99.7 GB4bit✗ too large198.9 GB8bit
cerebras/Qwen3-Coder-REAP-25B-A3BOpen weightsCerebras SystemsApache-2.024.9B3B active20 Oct 20250✓ fits14.8 GB4bit✓ fits29.1 GB8bit
chatglm2-6bOpen weightsZ.ai (Zhipu AI)unclassified6B24 Jun 20230✓ fits4.0 GB4bit✓ fits7.4 GB8bit
chatglm3-6bOpen weightsZ.ai (Zhipu AI)unclassified6.24B25 Oct 20230✓ fits4.1 GB4bit✓ fits7.7 GB8bit
chronos-2Open weightsAmazon Web ServicesApache-2.0119.5M30 Oct 20250✓ fits0.6 GB4bit✓ fits0.6 GB8bit
chronos-bolt-baseOpen weightsAmazon Web ServicesApache-2.0205.3M25 Nov 20240✓ fits0.6 GB4bit✓ fits0.7 GB8bit
chronos-bolt-miniOpen weightsAmazon Web ServicesApache-2.021.2M25 Nov 20240✓ fits0.5 GB4bit✓ fits0.5 GB8bit
chronos-bolt-smallOpen weightsAmazon Web ServicesApache-2.047.7M25 Nov 20240✓ fits0.5 GB4bit✓ fits0.6 GB8bit
chronos-bolt-tinyOpen weightsAmazon Web ServicesApache-2.08.65M25 Nov 20240✓ fits0.5 GB4bit✓ fits0.5 GB8bit
chronos-t5-baseOpen weightsAmazon Web ServicesApache-2.0201.4M21 Feb 20240✓ fits0.6 GB4bit✓ fits0.7 GB8bit
chronos-t5-largeOpen weightsAmazon Web ServicesApache-2.0709M21 Feb 20240✓ fits0.9 GB4bit✓ fits1.3 GB8bit
chronos-t5-miniOpen weightsAmazon Web ServicesApache-2.020.5M21 Feb 20240✓ fits0.5 GB4bit✓ fits0.5 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