353K
Soliloquy-L3 is a fast, highly capable roleplaying model designed for immersive, dynamic experiences. Trained on over 250 million tokens of roleplaying data, Soliloquy-L3 has a vast knowledge base, rich literary expression, and support for up to 24k context length. It outperforms existing ~13B models, delivering enhanced roleplaying capabilities. Usage of this model is subject to [Meta's Acceptable Use Policy](https://ai.meta.com/llama/use-policy/). *This provider may log and train based on your prompt.*
185K
The Claude 3 Opus from Anthropic is their most advanced model, designed for highly complex tasks. It excels in performance, intelligence, fluency, and comprehension.
91K
Phi-3 Mini is a 3.8B parameters, lightweight, state-of-the-art open model trained with the Phi-3 datasets that includes both synthetic data and the filtered publicly available websites data with a focus on high-quality and reasoning dense properties. The model has underwent a post-training process that incorporates both supervised fine-tuning and direct preference optimization to ensure precise instruction adherence and robust safety measures. When assessed against benchmarks testing common sense, language understanding, math, code, long context and logical reasoning, Phi-3 Mini-4K-Instruct showcased a robust and state-of-the-art performance among models with less than 13 billion parameters. *This model and the API is experimental.*
89K
WizardLM-2 8x22B is Microsoft AI's most advanced Wizard model. It demonstrates highly competitive performance compared to those leading proprietary models.
85K
The Claude 3 Haiku, created by Anthropic, is their fastest and most efficient model to date, designed for near-instant response. It offers rapid and accurate performance for specific tasks.
77K
Mixtral 8x22B is mistral's latest open model. It sets a new standard for performance and efficiency within the AI community. It is a sparse Mixture-of-Experts (SMoE) model that uses only 39B active parameters out of 141B, offering unparalleled cost efficiency for its size.
71K
The Claude 3 Sonnet offers an optimal mix of intelligence and speed suitable for enterprise tasks. It provides high utility at a reduced cost, is reliable, and well-suited for large-scale deployments.
57K
An affordable 7B-parameter model that combines multiple models using the new task_arithmetic merge method from MergeKit.
57K
Meta developed and released the Meta Llama 3 family of large language models (LLMs), a collection of pretrained and instruction tuned generative text models in 8 and 70B sizes. The Llama 3 instruction tuned models are optimized for dialogue use cases and outperform many of the available open source chat models on common industry benchmarks.
53K
Meta developed and released the Meta Llama 3 family of large language models (LLMs), a collection of pretrained and instruction tuned generative text models in 8 and 70B sizes. The Llama 3 instruction tuned models are optimized for dialogue use cases and outperform many of the available open source chat models on common industry benchmarks.
48K
A compact LLM offering superior performance to GPT-3.5, with robust multilingual capabilities for both English and Korean, delivering high efficiency in a smaller package.
46K
Nous Hermes 2 Mixtral 8x7B DPO is the new flagship Nous Research model trained over the Mixtral 8x7B MoE LLM.
41K
Zephyr 141B-A35B is an instruction finetuned version of Mixtral-8x22B. It was fine-tuned on a mix of publicly available, synthetic datasets trained by Huggingface. It achieves strong performance on chat benchmarks.
36K
Mistral-Tiny is the smallest and most cost-effective model in the lineup offered by Mistral AI, designed to provide essential language modeling capabilities in a resource-efficient package.
34K
Mistral AI's premier model. Built on a closed-source prototype, it is highly capable in reasoning, coding, handling JSON, chatting, and more. For more details, refer to the launch announcement. The model is proficient in English, French, Spanish, German, and Italian, delivering high grammatical precision. With a context window of 32K tokens, it effectively recalls detailed information from extensive documents.
24K
Mistral Medium is a closed-source model developed by Mistral AI. It operates on a proprietary model weights and excels in reasoning, coding, handling JSON, chatting, and various other applications. It performs comparably to many flagship models from different companies in benchmarks.
23K
通义千问-72B(Qwen-72B)是阿里云研发的通义千问大模型系列的720亿参数规模的模型。Qwen-72B是基于Transformer的大语言模型, 在超大规模的预训练数据上进行训练得到。预训练数据类型多样,覆盖广泛,包括大量网络文本、专业书籍、代码等。 Qwen-72B (Qwen-72B) is a 72 billion parameter scale model of the Qwen Big Model series developed by Alibaba Cloud which is trained on super-large scale pre-training data. The pre-training data are of various types and cover a wide range, including a large number of web texts, professional books, codes, etc.
18K
Mistral-Small is a mid-tier model in the Mistral AI suite, offering a balance between performance and affordability.
15K
Context-aware English-Korean translation that leverages previous dialogues to ensure unmatched coherence and continuity in your conversations.
12K
Command is a conversational model designed for following instructions and executing language tasks with superior quality, greater reliability, and a broader context compared to our standard generative models. Utilization of this model is governed by Cohere’s Acceptable Use Policy.
10K
Arctic is a dense-MoE Hybrid transformer architecture pre-trained from scratch by the Snowflake AI Research Team. **Efficient Intelligence**: Arctic outperforms similar open source models in enterprise tasks like SQL generation and coding, setting a new standard for cost-effective AI training for Snowflake customers. **True Openness**: Licensed under Apache 2.0, Arctic offers full access to its code and weights, and openly shares all data recipes and research insights.
5K
Qwen1.5 is the improved version of Qwen, the large language model series developed by Qwen team, Alibaba Cloud.
5K
Sonar represents the newest model family from Perplexity, offering improvements over previous models in terms of cost-efficiency, speed, and performance.
5K
Gemini 1.5 by Google delivers dramatically enhanced performance with a more efficient architecture. This model is intended for early testing.
3K
GPT-4 is a large multimodal model (accepting text or image inputs and outputting text) that can solve difficult problems with greater accuracy than any of our previous models, thanks to its broader general knowledge and advanced reasoning capabilities.
3K
Phind-CodeLlama-34B-v2 is an open-source language model, fine-tuned on 1.5B tokens from high-quality programming-related data, and proficient in languages like Python, C/C++, TypeScript, and Java. It achieved a 73.8% pass rate on HumanEval and is instruction-tuned using Alpaca/Vicuna formats for better usability and steerability. Trained on proprietary instruction-answer pairs, it generates a single completion per prompt.
3K
Sonar represents the newest model family from Perplexity, offering improvements over previous models in terms of cost-efficiency, speed, and performance.
2K
Sonar represents the newest model family from Perplexity, offering improvements over previous models in terms of cost-efficiency, speed, and performance. For *-online* models, in addition to the token charges, a flat $5 is charged per thousand requests (or half a cent per request).
1K
Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models. They are text-to-text, decoder-only large language models, available in English, with open weights, pre-trained variants, and instruction-tuned variants. Gemma models are well-suited for a variety of text generation tasks, including question answering, summarization, and reasoning. Their relatively small size makes it possible to deploy them in environments with limited resources such as a laptop, desktop or your own cloud infrastructure, democratizing access to state of the art AI models and helping foster innovation for everyone.
933
Airoboros is a fairly general purpose model, but focuses heavily on instruction following, rather than casual chat/roleplay.
833
Mythomax L2 13B is a large language model created by Gryphe that specializes in storytelling and advanced roleplaying. It is built on the Llama 2 architecture and is an optimized version of the MythoMix model, incorporating a tensor merger strategy for increased coherency and performance. Mythomax L2 13B has been praised for its ability to bond characters and create engaging roleplaying experiences.
555
Vision language model LLaVA 1.6 allowing both image and text inputs. LLaVA is an open-source chatbot trained by fine-tuning LLM on multimodal instruction-following data.
502
The latest GPT-3.5 Turbo model with higher accuracy at responding in requested formats and a fix for a bug which caused a text encoding issue for non-English language function calls. Returns a maximum of 4,096 output tokens.
370
Mixtral is mixture of expert large language model (LLM) from Mistral AI. This is state of the art machine learning model using a mixture 8 of experts (MoE) 7b models. During inference 2 expers are selected. This architecture allows large models to be fast and cheap at inference. The Mixtral-8x7B outperforms Llama 2 70B on most benchmarks.
325
Hermes 2 Pro is an upgraded, retrained version of Nous Hermes 2, consisting of an updated and cleaned version of the OpenHermes 2.5 Dataset, as well as a newly introduced Function Calling and JSON Mode dataset developed in-house.
252
Qwen1.5 is the improved version of Qwen, the large language model series developed by Qwen team, Alibaba Cloud.
202
OpenHermes 2 Mistral 7B is a state of the art Mistral Fine-tune.
178
Qwen1.5 is the improved version of Qwen, the large language model series developed by Qwen team, Alibaba Cloud.
174
Qwen1.5 is the improved version of Qwen, the large language model series developed by Qwen team, Alibaba Cloud.
150
Llama 2 is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. This is 70B fine-tuned model, optimized for dialogue use cases and converted for the Hugging Face Transformers format.
121
This model is a 75/25 merge of Chronos (13B) and Nous Hermes (13B) models resulting in having a great ability to produce evocative storywriting and follow a narrative.
101
SOLAR-10.7B is a large language model with 10.7 billion parameters, showing superior performance across various natural language processing tasks, outperforming other models with up to 30 billion parameters. This model employs depth up-scaling for enhancement, integrating architectural changes and continuing pretraining with Mistral 7B weights. It excels in robustness and adaptability, making it ideal for fine-tuning applications, and consistently surpasses the Mixtral 8X7B model in benchmarks.
68
Nous Hermes 2 - Yi-34B is a state of the art Yi Fine-tune. Nous Hermes 2 Yi 34B was trained on 1,000,000 entries of primarily GPT-4 generated data
36
Yi-34B is a large language model (LLM) developed by the AI startup 01.AI, It is a bilingual (English and Chinese) base model trained with 34 billion parameters. Yi-34B has shown impressive performance on various natural language processing tasks.
The Capybara series is the first Nous collection of dataset and models made by fine-tuning mostly on data created by Nous in-house.
OpenHermes 2.5 Mistral 7B is a state of the art Mistral Fine-tune, a continuation of OpenHermes 2 model, which trained on additional code datasets.
OLMo is a series of **O**pen **L**anguage **Mo**dels designed to enable the science of language models. The OLMo base models are trained on the [Dolma](https://huggingface.co/datasets/allenai/dolma) dataset. The adapted versions are trained on the [Tulu SFT mixture](https://huggingface.co/datasets/allenai/tulu-v2-sft-mixture) and, for the Instruct version, a [cleaned version of the UltraFeedback dataset](https://huggingface.co/datasets/allenai/ultrafeedback_binarized_cleaned). OLMo 7B Instruct and OLMo SFT are two adapted versions of these models trained for better question answering. They show the performance gain that OLMo base models can achieve with existing fine-tuning techniques.
Sonar represents the newest model family from Perplexity, offering improvements over previous models in terms of cost-efficiency, speed, and performance. For *-online* models, in addition to the token charges, a flat $5 is charged per thousand requests (or half a cent per request).
Qwen1.5 is the improved version of Qwen, the large language model series developed by Qwen team, Alibaba Cloud.
Qwen1.5 is the improved version of Qwen, the large language model series developed by Qwen team, Alibaba Cloud.
Qwen1.5 is the improved version of Qwen, the large language model series developed by Qwen team, Alibaba Cloud.
LLaVA vision-language model trained on OSS LLM generated instruction following data. 1 Image is counted as 576 prompt tokens.
This is a 7B-parameter decoder-only Japanese language model fine-tuned on instruction-following datasets, built on top of the base model Japanese Stable LM Base Gamma 7B.
japanese-stablelm-base-beta-70b is a 70B-parameter decoder-only language model based on Llama-2-70b that has been fine-tuned on a diverse collection of Japanese data, with the intent of maximizing downstream performance on Japanese language tasks.
DBRX Instruct is a mixture-of-experts (MoE) large language model trained from scratch by Databricks. DBRX Instruct specializes in few-turn interactions.
OpenChat is a library of open-source language models that have been fine-tuned with C-RLFT, a strategy inspired by offline reinforcement learning. These models can learn from mixed-quality data without preference labels and have achieved exceptional performance comparable to ChatGPT.
A Mythomax/MLewd_13B-style merge of selected 70B models A multi-model merge of several LLaMA2 70B finetunes for roleplaying and creative work. The goal was to create a model that combines creativity with intelligence for an enhanced experience.
WizardLM-2 7B is the smaller variant of Microsoft AI's latest Wizard model. It is the fastest and achieves comparable performance with existing 10x larger open-source leading models
Context-aware Korean-English translation that leverages previous dialogues to ensure unmatched coherence and continuity in your conversations.
The Mistral-7B-Instruct-v0.2 Large Language Model (LLM) is an enhanced version of the Mistral-7B-v0.2 generative text model, fine-tuned for instruction-based tasks using numerous publicly accessible conversation datasets.