Dolphin Llama 3 8B
Dolphin Llama 3 8B is the compact counterpart to Dolphin Llama 3 70B, applying the same uncensored fine-tuning approach to Meta's Llama 3 8B base. It targets deployments where GPU memory is limited but developers still require unfiltered instruction-following behavior for their specific application context.
At 8B parameters the model offers fast inference on mid-range consumer hardware while retaining reasonable general capability for its size. Like the 70B variant, it is aimed at developers building systems with their own content governance layer. Its lower resource requirements make it a practical choice for local or self-hosted deployments in research and development environments.
Key Features
Uncensored instruction following on a compact 8B footprint
Fast inference on consumer GPUs and mid-range hardware
General-purpose text generation without upstream refusal filters
System-prompt-driven behavior customization
Multi-turn conversational capability
Self-hostable for privacy-sensitive or offline use cases
Ideal Use Cases
Local uncensored assistant for researcher and developer use
Privacy-focused self-hosted chatbot deployment
Rapid prototyping of applications requiring flexible content policies
Lightweight red-teaming and adversarial testing tool
Example Prompts for Dolphin Llama 3 8B
Technical Specifications
| Provider | Dolphin |
| Category | Text |
| Modality | Text -> Text |
Frequently Asked Questions
Try Dolphin Llama 3 8B now
Start using Dolphin Llama 3 8B instantly — 100 free credits, no credit card required. Access 750+ AI models through one platform.