Downloads 30d

30.9M

Hub likes

1.8K

Params

752M

The original GPT-2 checkpoint. Still the most downloaded causal language model on the Hub and a standard baseline.

Downloads 30d

15.4M

Hub likes

4.2K

Params

137M

Mid-size dense model from the Qwen3 family, with switchable thinking and non-thinking inference modes.

Downloads 30d

10M

Hub likes

2.1K

Params

8.2B

Downloads 30d

5.6M

Hub likes

1.1K

Params

31B

Instruction-tuned 7B chat model with strong multilingual coverage and a permissive licence.

Downloads 30d

8.1M

Hub likes

2.4K

Params

7.6B

Downloads 30d

5M

Hub likes

644

Params

18.7B

Downloads 30d

7.6M

Hub likes

663

Params

494M

Smallest checkpoint in Meta’s Open Pre-trained Transformer suite, released to reproduce GPT-3 class training openly.

Downloads 30d

6.1M

Hub likes

324

Params

125M

Downloads 30d

1.6M

Hub likes

22

Params

31.3B

Small instruction-tuned model that runs comfortably on modest hardware.

Downloads 30d

6.6M

Hub likes

870

Params

1.5B

Small Qwen3 model supporting both thinking and direct-answer modes in a single checkpoint.

Downloads 30d

9.6M

Hub likes

730

Params

4B

Compact instruction-tuned model built for on-device assistants and summarisation at low latency.

Downloads 30d

7.2M

Hub likes

1.8K

Params

1.2B

OpenAI’s smaller open-weight mixture-of-experts release, roughly 21B total parameters with about 3.6B active per token.

Downloads 30d

6M

Hub likes

5.1K

Params

20.9B

The most-liked chat model on the Hub. 128K context, instruction-tuned, widely used as a fine-tuning base.

Downloads 30d

6.1M

Hub likes

8.3K

Params

8B

Downloads 30d

4.3M

Hub likes

592

Params

3.1B

Downloads 30d

5.2M

Hub likes

511

Params

9B

The larger open-weight OpenAI release, around 117B total parameters with roughly 5.1B active per token.

Downloads 30d

4M

Hub likes

5.4K

Params

116.8B

Largest dense model in the Qwen3 line, aimed at reasoning and long-context work under a permissive licence.

Downloads 30d

3.6M

Hub likes

754

Params

32.8B

Downloads 30d

4.7M

Hub likes

69

Params

34.4B

Downloads 30d

4.5M

Hub likes

4K

Params

304.2B

Downloads 30d

3.7M

Hub likes

494

Params

35B

Downloads 30d

3.6M

Hub likes

997

Params

4B

Downloads 30d

887.1K

Hub likes

364

Params

72.3B

Lightweight Qwen3 checkpoint aimed at latency-sensitive chat deployments.

Downloads 30d

3.3M

Hub likes

568

Params

2B

Downloads 30d

2.5M

Hub likes

53

Params

7.6B

Downloads 30d

2.6M

Hub likes

675

Params

9B

Part of a model suite trained on identical data in identical order, built specifically for interpretability research.

Downloads 30d

3.5M

Hub likes

46

Params

213M

Downloads 30d

1.6M

Hub likes

130

Params

4B

Downloads 30d

6.1M

Hub likes

58

Params

2.2B

Instruction-tuned small model from the Gemma 3 family, built from the same research as Gemini.

Downloads 30d

3M

Hub likes

1.2K

Params

1B

Code-specialised instruction model covering generation, completion and repair across many languages.

Downloads 30d

2M

Hub likes

817

Params

7.6B

Downloads 30d

2.4M

Hub likes

1.5K

Params

27B

Downloads 30d

3M

Hub likes

75

Params

14.8B

Mid-size instruction-tuned chat model, a common step up when 7B is not quite capable enough.

Downloads 30d

1.3M

Hub likes

371

Params

14.8B

Downloads 30d

1.8M

Hub likes

362

Params

32.8B

Downloads 30d

4.4M

Hub likes

1.5K

Params

685.4B

Downloads 30d

1.7M

Hub likes

1.1K

Params

35B

Very small language model trained on a curated corpus, intended for on-device and edge deployment.

Downloads 30d

1.7M

Hub likes

246

Params

135M

Downloads 30d

1.5M

Hub likes

506

Params

1M

Downloads 30d

2.1M

Hub likes

649

Params

88M

Downloads 30d

1.7M

Hub likes

39

Params

14.8B

Downloads 30d

1.3M

Hub likes

478

Downloads 30d

1.3M

Hub likes

645

Params

27B

Downloads 30d

1.5M

Hub likes

1.9K

Params

31.2B

Downloads 30d

876.5K

Hub likes

157

Params

14.4B

Mid-size member of the Qwen2.5-Coder family, trained on a large code-heavy corpus.

Downloads 30d

1.6M

Hub likes

193

Params

14.8B

Mixture-of-experts model with roughly 30B total parameters and about 3B active, trading memory for inference speed.

Downloads 30d

1.4M

Hub likes

956

Params

30.5B

Downloads 30d

1.5M

Hub likes

2.8K

Params

3.2B

Downloads 30d

1.6M

Hub likes

3.2K

Params

7.2B

Downloads 30d

2.3M

Hub likes

55

Params

8.2B

Downloads 30d

932.8K

Hub likes

2.3K

Params

290.9B

Downloads 30d

1.4M

Hub likes

578

Params

20.9B

Downloads 30d

1.4M

Hub likes

464

Params

494M

Downloads 30d

341.8K

Hub likes

0

Downloads 30d

5.1M

Hub likes

491

Params

14.8B

Downloads 30d

609.2K

Hub likes

84

Params

1.7B

Downloads 30d

596.5K

Hub likes

101

Params

4B

Downloads 30d

1M

Hub likes

194

Params

79.7B

Downloads 30d

963.1K

Hub likes

42

Params

32.8B

Downloads 30d

1.7M

Hub likes

433

Params

135M