LLM ReferenceLLM Reference
10 models2023Up to 32K ctxFrom $0.2/1M input

About

The OpenOrca family of large language models (LLMs) leverages the Mistral 7B base model, fine-tuned with the OpenOrca dataset. This comprehensive dataset, mirroring the Orca Research Paper dataset, includes approximately 1 million GPT-4 completions and 3.2 million GPT-3.5 completions. OpenOrca models excel in a range of natural language processing tasks such as text and code generation, question answering, and conversations. Among these, the Mistral-7B-OpenOrca stands out for its superior performance among models with less than 30 billion parameters, achieving 98% of Llama2-70B-chat's performance on the HuggingFace leaderboard at its release. Designed to run efficiently on consumer-grade GPUs, these models are valuable tools for developers, though users should be mindful of potential biases inherent to the LLMs' training data1246810.

Specifications(10 models)

OpenOrca model specifications comparison
ModelReleasedContextParameters
Mixtral 8x7B SlimOrca2023-1232K56B
Mistral 7B SlimOrca2023-108K7B
Mistral 7B OpenOrca2023-108K7B
Mistral 7B OpenOrca Platypus2023-108K7B
OpenOrca Phi 1.52023-094K1.5B
LlongOrca 13B 16K2023-0816K13B
OpenOrca Platypus2 13B2023-084K13B
LlongOrca 7B 16K2023-0816K7B
OpenOrca x OpenChat 13B2023-074K13B
OpenOrca 13B2023-074K13B

Available From(2 providers)

Pricing

OpenOrca model pricing by provider
ModelProviderInput / 1MOutput / 1MType
Mistral 7B OpenOrcaTogether AI$0.2$0.2Serverless
Mistral 7B OpenOrcaFireworks AI$0.2$0.2Provisioned

Frequently Asked Questions

What is OpenOrca?
The OpenOrca family of large language models (LLMs) leverages the Mistral 7B base model, fine-tuned with the OpenOrca dataset. This comprehensive dataset, mirroring the Orca Research Paper dataset, includes approximately 1 million GPT-4 completions and 3.2 million GPT-3.5 completions. OpenOrca models excel in a range of natural language processing tasks such as text and code generation, question answering, and conversations. Among these, the Mistral-7B-OpenOrca stands out for its superior performance among models with less than 30 billion parameters, achieving 98% of Llama2-70B-chat's performance on the HuggingFace leaderboard at its release. Designed to run efficiently on consumer-grade GPUs, these models are valuable tools for developers, though users should be mindful of potential biases inherent to the LLMs' training data1246810.
How many models are in the OpenOrca family?
The OpenOrca family contains 10 models.
What is the latest OpenOrca model?
The latest model is Mixtral 8x7B SlimOrca, released in 2023-12.
How much does OpenOrca cost?
OpenOrca models are available at $0.2/1M input tokens through providers like Together AI.

Models(10)