Explore Our Models

Qrwkv-72B is a linear-attention RWKV variant of the Qwen 2.5 72B model, optimized to significantly reduce computational cost at scale. Leveraging linear attention, it achieves substantial inference speedups (>1000x) while retaining competitive accuracy on common benchmarks like ARC, HellaSwag, Lambada, and MMLU. It inherits knowledge and language support from Qwen 2.5, supporting approximately 30 languages, making it suitable for efficient inference in large-context applications.

Eagle 7B is trained on 1.1 Trillion Tokens across 100+ world languages (70% English, 15% multilang, 15% code). - Built on the [RWKV-v5](/models?q=rwkv) architecture (a linear transformer with 10-100x+ lower inference cost) - Ranks as the world's greenest 7B model (per token) - Outperforms all 7B class models in multi-lingual benchmarks - Approaches Falcon (1.5T), LLaMA2 (2T), Mistral (>2T?) level of performance in English evals - Trade blows with MPT-7B (1T) in English evals - All while being an ["Attention-Free Transformer"](https://www.isattentionallyouneed.com/) Eagle 7B models are provided for free, by [Recursal.AI](https://recursal.ai), for the beta period till end of March 2024 Find out more [here](https://blog.rwkv.com/p/eagle-7b-soaring-past-transformers) [rnn](/models?q=rwkv)

[RWKV](https://wiki.rwkv.com) is an RNN (recurrent neural network) with transformer-level performance. It aims to combine the best of RNNs and transformers - great performance, fast inference, low VRAM, fast training, "infinite" context length, and free sentence embedding. RWKV-5 is trained on 100+ world languages (70% English, 15% multilang, 15% code). RWKV 3B models are provided for free, by Recursal.AI, for the beta period. More details [here](https://substack.recursal.ai/p/public-rwkv-3b-model-via-openrouter). #rnn

This is an [RWKV 3B model](/models/rwkv/rwkv-5-world-3b) finetuned specifically for the [AI Town](https://github.com/a16z-infra/ai-town) project. [RWKV](https://wiki.rwkv.com) is an RNN (recurrent neural network) with transformer-level performance. It aims to combine the best of RNNs and transformers - great performance, fast inference, low VRAM, fast training, "infinite" context length, and free sentence embedding. RWKV 3B models are provided for free, by Recursal.AI, for the beta period. More details [here](https://substack.recursal.ai/p/public-rwkv-3b-model-via-openrouter). #rnn