Downloads 30d

26.7M

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

18.6M

Hub likes

3.9K

Params

137M

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

Downloads 30d

8.3M

Hub likes

2.1K

Params

8.2B

Downloads 30d

461.5K

Hub likes

1.2K

Params

31B

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

Downloads 30d

6.6M

Hub likes

2.5K

Params

7.6B

Downloads 30d

4.2M

Hub likes

670

Params

18.7B

Downloads 30d

5.5M

Hub likes

709

Params

494M

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

Downloads 30d

5.8M

Hub likes

331

Params

125M

Downloads 30d

55.5K

Hub likes

30

Params

31.3B

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

Downloads 30d

4.9M

Hub likes

905

Params

1.5B

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

Downloads 30d

11.6M

Hub likes

769

Params

4B

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

Downloads 30d

4.3M

Hub likes

1.9K

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

5.8M

Hub likes

5.2K

Params

20.9B

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

Downloads 30d

5M

Hub likes

8.5K

Params

8B

Downloads 30d

3.9M

Hub likes

615

Params

3.1B

Downloads 30d

5.1M

Hub likes

573

Params

9B

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

Downloads 30d

3M

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

2.8M

Hub likes

771

Params

32.8B

Downloads 30d

4.6M

Hub likes

85

Params

34.4B

Downloads 30d

4.3M

Hub likes

4.1K

Params

304.2B

Downloads 30d

3.9M

Hub likes

540

Params

35B

Downloads 30d

2.7M

Hub likes

1K

Params

4B

Downloads 30d

5.1K

Hub likes

368

Params

72.3B

Lightweight Qwen3 checkpoint aimed at latency-sensitive chat deployments.

Downloads 30d

4.2M

Hub likes

591

Params

2B

Downloads 30d

2M

Hub likes

59

Params

7.6B

Downloads 30d

2.4M

Hub likes

695

Params

9B

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

Downloads 30d

3.4M

Hub likes

62

Params

213M

Downloads 30d

2.5M

Hub likes

139

Params

4B

Downloads 30d

3.9M

Hub likes

76

Params

2.2B

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

Downloads 30d

1.2M

Hub likes

1.3K

Params

1B

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

Downloads 30d

1.5M

Hub likes

841

Params

7.6B

Downloads 30d

2.5M

Hub likes

1.7K

Params

27B

Downloads 30d

2.6M

Hub likes

87

Params

14.8B

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

Downloads 30d

826.5K

Hub likes

377

Params

14.8B

Downloads 30d

1.3M

Hub likes

372

Params

32.8B

Downloads 30d

5.5M

Hub likes

1.6K

Params

685.4B

Downloads 30d

1.5M

Hub likes

1.1K

Params

35B

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

Downloads 30d

1.8M

Hub likes

259

Params

135M

Downloads 30d

1.4M

Hub likes

518

Params

1M

Downloads 30d

1.8M

Hub likes

658

Params

88M

Downloads 30d

1.4M

Hub likes

51

Params

14.8B

Downloads 30d

1.2M

Hub likes

490

Downloads 30d

1M

Hub likes

728

Params

27B

Downloads 30d

1M

Hub likes

1.9K

Params

31.2B

Downloads 30d

1M

Hub likes

165

Params

14.4B

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

Downloads 30d

1.7M

Hub likes

203

Params

14.8B

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

Downloads 30d

1M

Hub likes

971

Params

30.5B

Downloads 30d

1.1M

Hub likes

2.8K

Params

3.2B

Downloads 30d

1.7M

Hub likes

3.3K

Params

7.2B

Downloads 30d

2.1M

Hub likes

67

Params

8.2B

Downloads 30d

802K

Hub likes

2.3K

Params

290.9B

Downloads 30d

921.9K

Hub likes

585

Params

20.9B

Downloads 30d

1.2M

Hub likes

482

Params

494M

Downloads 30d

3K

Hub likes

0

Downloads 30d

9.2M

Hub likes

513

Params

14.8B

Downloads 30d

555.4K

Hub likes

94

Params

1.7B

Downloads 30d

587.3K

Hub likes

114

Params

4B

Downloads 30d

990.2K

Hub likes

206

Params

79.7B

Downloads 30d

986.7K

Hub likes

48

Params

32.8B

Downloads 30d

1.2M

Hub likes

458

Params

135M