Supported Models
Text, Image, and Audio models currently available on Hyperbolic
| Category | Model | Precision / Description |
|---|---|---|
| Text | QwQ-32B | BF16 |
| Text | DeepSeek-R1 | FP8 |
| Text | DeepSeek-V3 | FP8 |
| Text | Llama-3-3-70B | FP8 |
| Text | QwQ-32B-Preview | FP8 |
| Text | Qwen2.5-Coder-32B | FP8 |
| Text | Llama-3.2-3B | FP8 |
| Text | Qwen2.5-72B | BF16 |
| Text | Llama-3-70B | FP8 |
| Text | Hermes-3-70B | FP8 |
| Text | Llama-3.1-405B | BF16 |
| Text | Llama-3.1-8B | FP8 |
| Text | Llama-3.1-70B | FP8 |
| Image | Flux.1 [dev] | SOTA prompt-following & visual quality |
| Image | SDXL-1.0 | High-resolution master |
| Image | Stable Diffusion-1.5 | Reliable classic |
| Image | Segmind SD-1B | Domain specialist (e.g. medical, scientific) |
| Image | Stable Diffusion-2 | Enhanced innovator |
| Image | SDXL-1.0-Turbo | Speed + high-res |
| ControlNet | SDXL-ControlNet | pose, edge, depth filters |
| ControlNet | SD1.5-ControlNet | canny, depth, openpose, softedge filters |
| Vision-Language | Qwen2.5-VL-72B-Instruct | BF16 |
| Vision-Language | Qwen2.5-VL-7B-Instruct | BF16 |
| Vision-Language | Pixtral-12B | BF16 |
| Base | Llama-3.1-405B-BASE (BF16) | BF16 |
| Audio | Melo TTS | Lifelike speech with smooth intonation |
Updated 7 days ago