Need to run open-source models or your own models at scale? We offer dedicated hosting.Schedule time with our team
Qwen3-Next-80B-A3B-Thinking
Qwen3-Next thinker.
gpt-oss-120b
OpenAI’s open-weight model (Big model smell)
DeepSeek-R1-0528
The latest open-source reasoner LLM released by DeepSeek.
Llama-3.1-405B-BASE
Still the SOTA base completion model but better because it's BF16.
![Flux.1 [dev]](data:image/gif;base64,R0lGODlhAQABAAD/ACwAAAAAAQABAAACADs%3D)
Flux.1 [dev]
A new SOTA image generation model that is outstanding in prompt following and visual quality.
All Models
Text Models
Qwen3-Next-80B-A3B-Thinking
Qwen3-Next thinker.
Qwen3-Next-80B-A3B-Instruct
Qwen3-Next Instruct model.
gpt-oss-120b
OpenAI’s open-weight model (Big model smell)
gpt-oss-20b
OpenAI’s open-weight model (Small model smell)
Qwen3-Coder-480B-A35B-Instruct
The latest and most powerful coder model from the Qwen Team.
Qwen3-235B-A22B-Instruct-2507
Qwen latest non-thinking model with significant improvements in general capabilities.
Qwen3-8B
A dense causal language model with Grouped Query Attention. Supports dual thinking modes for adaptive reasoning and general dialogue and optimized for multilingual capabilities.
Kimi-K2
Kimi's latest 1T LLM, good at coding and tool-calling.
DeepSeek-R1-0528
The latest open-source reasoner LLM released by DeepSeek.
DeepSeek-V3-0324
DeepSeek's updated V3 model released on 03/24/2025.
DeepSeek-R1
The best open-source reasoner LLM released by DeepSeek.
DeepSeek-V3
The best open-source LLM released by DeepSeek.
Llama-3.3-70B
Meta's latest 70B LLM with performance comparable to llama 3.1 405B
Qwen2.5-Coder-32B
The best coder from the Qwen Team.
Llama-3.2-3B
The latest Llama 3.2 instruction-tuned model by Meta.
Qwen2.5-72B
The latest Qwen LLM with more knowledge in coding and math.
Llama-3.1-405B
The Biggest and Best open-source AI model trained by Meta, beating GPT-4o across most benchmarks.
Llama-3.1-8B
The smallest and fastest member of the Llama 3.1 family.
Llama-3.1-70B
The best LLM at its size with faster response times compared to the 405B model.
Image Models
![Flux.1 [dev]](/_next/image?url=https%3A%2F%2Fa-us.storyblok.com%2Ff%2F1019538%2F1024x1024%2Fc1ead5b957%2Fflux-1-dev.png&w=750&q=75)
Flux.1 [dev]
A new SOTA image generation model that is outstanding in prompt following and visual quality.
SDXL-1.0
"The High-Resolution Master", excel at generating high-quality, detailed images with a focus on precision.
Stable Diffusion-1.5
"The Reliable Classic", known for its balanced performance and versatility.
Segmind SD-1B
"The Domain Specialist", optimized for niche applications, such as medical imaging, scientific visualization, etc.
Stable Diffusion-2
"The Enhanced Innovator", offers improved performance and capabilities over earlier versions.
SDXL-1.0-Turbo
"The Speedy Pro", combines high-resolution outputs with faster processing times.
VLM Models
Qwen/Qwen3-VL-32B-Thinking
Qwen3-VL-32B-Thinking is a powerful vision-language model that excels in text understanding, visual perception, and reasoning, making it suitable for complex tasks involving both text and images. It features enhanced capabilities for long-context processing, spatial reasoning, and multimodal interactions, allowing developers to create advanced applications that require deep integration of visual and textual data.
NVIDIA-Nemotron-Nano-12B-v2-VL-BF16
12B multimodal reasoning model for video understanding and document intelligence.
Qwen2.5-VL-72B-Instruct
Latest and biggest vision model from the Qwen Team.
Qwen2.5-VL-7B-Instruct
Vision model from the Qwen Team.
Pixtral-12B
Vision model from MistralAI.
