Capybara Models by Nous Research
About
The Capybara family of large language models (LLMs), developed by Nous Research, is known for its cutting-edge training techniques and robust performance. These models excel in bilingual tasks, particularly in English and Chinese, thanks to their Yi-34B model foundation 5. A standout feature is their remarkable ability to handle extensive contexts of up to 200,000 tokens 25, which facilitates a profound understanding of complex topics and nuanced dialogues. The training approach incorporates an innovative synthesis method known as Amplify-Instruct, leveraging diverse techniques and instructional data from datasets like Airoboros, EverythingLM, and Know_Logic, as well as LessWrong posts 235. This training results in models capable of sophisticated reasoning, engaging in multi-turn dialogues, and effectively summarizing intricate subjects 235. The family encompasses models with parameter sizes of 3B, 7B, and 34B, with larger variants planned 23. Additionally, a multimodal extension named Obsidian enhances their range of functionalities 2.
Current Variants
Use-when guidance is derived from seed capabilities, context, release, and replacement fields.
Use when the workload needs 200k context and 34B parameters.
Use when the workload needs 7B parameters and structured outputs.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Nous Capybara 34B | Use when the workload needs 200k context and 34B parameters. | 2024-10 | 200k context34B parameters | Current |
| Nous Capybara 7B | Use when the workload needs 7B parameters. | 2024-10 | 7B parameters | Current |
| Nous Capybara 7B V1.9 | Use when the workload needs 7B parameters and structured outputs. | 2024-10 | 7B parametersstructured outputs | Current |
| Nous Capybara 3B V1.9 | Use when the workload needs 3B parameters. | 2024-10 | 3B parameters | Current |
Release Timeline
1 release groupSpecifications(4 models)
| Model | Released | Context | Parameters | Structured Outputs |
|---|---|---|---|---|
| Nous Capybara 34B | 2024-10 | 200k | 34B | No |
| Nous Capybara 7B | 2024-10 | — | 7B | No |
| Nous Capybara 7B V1.9 | 2024-10 | — | 7B | Yes |
| Nous Capybara 3B V1.9 | 2024-10 | — | 3B | No |
Available From(2 providers)
Pricing
| Model | Provider | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|
| Nous Capybara 7B V1.9 | Fireworks AI | $0.2 | $0.2 | Provisioned |
| Nous Capybara 7B V1.9 | Together AI | $0.2 | $0.2 | Serverless |
| Nous Capybara 34B | Fireworks AI | $0.9 | $0.9 | Provisioned |
Frequently Asked Questions
- What is Capybara used for?
- Capybara is used for structured outputs and chatbot and role-playing use cases. The family description and listed model capabilities point to those workloads as the best fit.
- How does Capybara compare to Hermes 2?
- Capybara by Nous Research is strongest where you need structured outputs, while Hermes 2 by Nous Research is the closest related family to check for agent workflows and tool use. Capybara has 4 listed variants and reaches up to 200k context, while Hermes 2 reaches up to 200k context, so compare the specs and pricing tables before choosing a production model.
- Which Capybara model should I use?
- For the lowest listed input price, start with Nous Capybara 7B V1.9 through Fireworks AI at $0.2/1M input tokens. For the most capable/latest local choice, evaluate Nous Capybara 7B V1.9 with structured outputs.
