Small / on-device models
Sub-10B-parameter models that run on consumer GPUs or even phones. Trade-off: lower raw quality, but private, offline, and instant.
5 models in this category. See also the “Best small / on-device AI models” ranking with editorial picks and FAQ.
| Model | Provider | Context | Input price / 1M | Tier |
|---|---|---|---|---|
| Llama 3.1 8B Instruct meta/llama-3.1-8b-instruct | Meta | 128K Long context (128K+) | Custom | Variable |
| GPT-4o mini openai/gpt-4o-mini | OpenAI | 128K Long context (128K+) | Custom | Variable |
| o3-mini openai/o3-mini | OpenAI | 200K Long context (128K+) | Custom | Variable |
| Jamba 1.5 Mini ai21/jamba-1.5-mini | AI21 | 256K Long context (128K+) | Custom | Variable |
| Mistral Small mistral-ai/mistral-small | Mistral AI | 128K Long context (128K+) | Custom | Variable |