Models
Currently supported providers: OpenAI, Google, Anthropic, Meta, and Mistral AI. See the Model Inclusion Policy for details on selection criteria.
| Model | Provider | Type | Input Modality | Output Modality | Context (tokens) | Deployment Option | Status | Release Date |
|---|---|---|---|---|---|---|---|---|
| GPT-4o mini Realtime | OpenAI | Multimodal | Text, Audio | Text, Audio | 16k | Hosted API only | Active | 2024-12-17 |
| Gemini 1.0 Ultra | Multimodal | Text, Image | Text | — | Hosted API only | Decommissioned | 2024-12-06 | |
| Gemini 1.0 Nano | LLM | Text | Text | — | On-device only | Decommissioned | 2024-12-06 | |
| PaliGemma 2 (10B) | Vision | Text, Image | Text | 8k | Hosted API only | Active | 2024-12-05 | |
| o1 | OpenAI | LLM | Text, Image | Text | 200k | Hosted API only | Active | 2024-12-05 |
| PaliGemma 2 (28B) | Vision | Text, Image | Text | 8k | Self-hosted only | Active | 2024-12-05 | |
| Mistral Large 2.1 | Mistral AI | LLM | Text | Text | 128k | Self-hosted + Hosted API | Active | 2024-11-18 |
| Pixtral Large | Mistral AI | Multimodal | Text, Image | Text | 128k | Hosted API only | Active | 2024-11-18 |
| Mistral Moderation | Mistral AI | Moderation | Text | Classification | 8k | Hosted API only | Active | 2024-11-06 |
| Claude 3.5 Haiku | Anthropic | Multimodal | Text, Image | Text | 200k | Hosted API only | Active | 2024-11-04 |