DeepSeek-R1 is an advanced large language model (LLM) that achieves performance comparable to OpenAI's o1 across math, code, and reasoning tasks.

DeepSeek R1 demonstrates emergent behaviors like chain-of-thought reasoning, self-verification, and error correction.

DeepSeek-V3 is a 671B parameter Mixture-of-Experts (MoE) language model with 37B active parameters per token

DeepSeek V3 is a state-of-the-art open-source large language model (LLM) with 671 billion parameters, of which 37 billion are activated per token.

Hermes 3 70B is the latest large language model in the Hermes series developed by NousResearch.

Hermes 3 8B is the latest iteration in the Hermes series of large language models, built on Llama 3.1 8B.

Llama 3.3 70B Instruct is a powerful large language model developed by Meta AI.

Llama 3.3 70B Instruct with Speculative Decoding is Meta's latest large language model optimized for efficient inference

Llama 3.3 70B Instruct Turbo is a state-of-the-art large language model released by Meta on December 6, 2024.

QwQ-32B-Preview is an experimental research model developed by the Qwen Team, focusing on advancing AI reasoning capabilities

QwQ-32B-Preview is an experimental research model developed by the Qwen Team at Alibaba, focused on advancing AI reasoning capabilities.

Qwen2.5-Coder-32B-Instruct is a state-of-the-art open-source large language model for code generation, reasoning, and fixing.

Qwen2.5-Coder-32B-Instruct is a state-of-the-art open-source large language model for code generation, reasoning, and fixing.

Llama 3.1 8B is a flexible model from Meta that is very capable for its size. It has a context length of 128k and can provide 8,192 output tokens.

Llama 3.2 1B is an extremely small, fast, and cheap model from Meta that is very capable for its size.

Ministral 8B is an 8 billion parameter large language model developed by Mistral AI, featuring a unique interleaved sliding-window attention pattern for faster, memory-efficient inference.

Llama 3.1 Nemotron 70B Instruct is a large language model developed by NVIDIA, based on Meta's Llama 3.1 70B architecture

Llama 3.2 11B Vision (Preview) is a powerful multimodal large language model developed by Meta

Llama 3.2 3B is a lightweight, text-only large language model developed by Meta. It features 3 billion parameters and a 131,072 token context window

Llama 3.2 90B Vision (Preview) is a large multimodal language model developed by Meta, designed for image-in, text-out tasks.

Llama 3.2 3B Instruct Turbo is a compact and efficient large language model developed by Meta AI

Gemini 1.5 Flash-8B is a compact and efficient large language model developed by Google DeepMind.

Qwen2.5-72B-Instruct is a large language model with 72 billion parameters, part of the Qwen2.5 series.

Qwen2.5-VL-72B-Instruct is a large vision-language model that combines advanced visual understanding with instruction-tuned language capabilities

Mistral Small 3 is a cutting-edge 24B-parameter language model released in January 2025 by Mistral AI

OpenAI's o1-mini is a smaller, faster, and more cost-effective version of the o1 model series, designed for complex reasoning tasks

OpenAI's o1-preview is an advanced large language model designed to excel at complex reasoning tasks.

Pixtral 12B is an advanced multimodal AI model developed by Mistral AI. It combines a 12-billion-parameter text decoder with a vision encoder, enabling it to process and understand both text and images simultaneously.

Hermes 3 405B is a frontier-level large language model developed by Nous Research, representing a full parameter finetune of the Llama-3.1 405B foundation model.

Llama 3.3 70B Instruct is the December update of Llama 3.1 70B and provides similar performance as 3.1 405B but with significant speed and cost improvements.

Llama 3.1 70B Instruct is a large language model developed by Meta with 70 billion parameters.

Meta Llama 3.1 8B Instruct Turbo is a multilingual large language model developed by Meta, it features a context window of 128k tokens.

Llama 3.1 405B Instruct is Meta's largest open-source language model, released on July 23, 2024

The Llama 3.1 8B Instruct model is an auto-regressive language model developed by Meta, featuring 8 billion parameters

Meta Llama 3.1 70B Instruct Turbo is a large language model developed by Meta, released on July 23, 2024

Meta Llama 3.1 405B Instruct Turbo is a state-of-the-art large language model developed by Meta.

Mistral NeMo is a 12 billion parameter large language model (LLM) developed collaboratively by Mistral AI and NVIDIA

GPT-4o Mini is OpenAI's latest cost-efficient small model, offering advanced AI capabilities at a fraction of the cost of larger models.

Gemma 2 9B Instruct is an open-source large language model developed by Google, released in 2024

Codestral 25.01 is Mistral AI's latest code generation model

Gemini 1.5 Pro is Google DeepMind's advanced multimodal large language model

Text Embedding 004 is a large language model designed for generating high-quality text embeddings.

Gemini 1.5 Flash is a Multimodal LLM designed for high-volume, rapid processing tasks. It excels in efficient handling of multimodal inputs (text, images, audio, and video) while maintaining a balance between performance and computational cost. Some other noteworthy features of Gemini 1.5 Flash include its ability to process up to 1 million tokens in context and its optimization for speed and efficiency in tasks such as summarization, chat, image and video captioning, and data extraction. | Metric | Value | |--------------------|--------------------| | Parameter Count | Unknown | | Mixture of Experts | Yes | | Context Length | 1,000,000 tokens | | Multilingual | Yes | | Quantized* | Unknown | \*_Quantization is specific to the inference provider and the model may be offered with different quantization levels by other providers._

GPT-4o is OpenAI's latest large language model, released in May 2024

Mixtral 8x22B is Mistral AI's most advanced open-source large language model, as of April 2024

Mistral Large is an LLM that excels in complex multilingual reasoning tasks, including text understanding, transformation, and code generation. It is particularly proficient in English, French, Spanish, German, and Italian, with a nuanced understanding of grammar and cultural context. Some other noteworthy features of Mistral Large include its ability to handle long-form content with a 32K token context window and its native support for function calling and JSON output. | Metric | Value | |--------------------|--------------------| | Parameter Count | 123 billion | | Mixture of Experts | No | | Context Length | 32,768 tokens | | Multilingual | Yes | | Quantized* | Unknown | \*_Quantization is specific to the inference provider and the model may be offered with different quantization levels by other providers._

Text Embedding 3 - Small is an Embeddings Model designed for generating high-quality vector representations of text. It excels in producing compact and meaningful embeddings for various natural language processing tasks, offering improved performance over its predecessor while maintaining efficiency in terms of latency and storage. Some other noteworthy features of Text Embedding 3 - Small include multilingual support and the ability to adjust embedding dimensions. | Metric | Value | |--------------------|--------------------| | Parameter Count | Unknown | | Mixture of Experts | No | | Context Length | Unknown | | Multilingual | Yes | | Quantized* | Unknown | \*_Quantization is specific to the inference provider and the model may be offered with different quantization levels by other providers._

Text Embedding 3 - Large is an advanced embedding model that converts text into high-dimensional vector representations.

Mixtral 8x7B is an open-source large language model developed by Mistral AI. It utilizes a Sparse Mixture of Experts (SMoE) architecture, allowing access to 47 billion total parameters while only using 13 billion active parameters per token.

Mistral 7B is an open-source large language model released by Mistral AI in September 2023

High quality edge model with only 3B parameters. Test this model through our UI, and then download the weights and run it on the edge on edge devices.