Explore the frontier of AI models in one place
Test 14 state-of-the-art language, vision, embedding, and image generation models through a clean interface or a REST API. No setup required.
DeepSeek V4 Pro
DeepSeek's flagship model excelling at multi-step reasoning, code generation, and complex analysis tasks. Ideal for software engineering and logical problem solving.
Try NowDeepSeek V4 Flash
Fast reasoning model that exposes its chain-of-thought process. Watch the model think step by step before delivering its answer.
Try NowNVIDIA Nemotron 550B
NVIDIA's flagship 550-billion-parameter model with extended reasoning budget. Exceptional for research, deep analysis, and expert-level tasks.
Try NowLlama 3.2 11B Vision
Meta's compact vision-language model capable of understanding images alongside text. Fast and efficient for visual question answering.
Try NowLlama 3.2 90B Vision
Meta's large 90-billion-parameter vision model. Superior image understanding and complex visual reasoning at scale.
Try NowPhi-4 Multimodal
Microsoft's efficient multimodal model delivering strong performance on text and vision tasks with a compact footprint.
Try NowPaliGemma
Google's open vision-language model pairing a SigLIP image encoder with a Gemma decoder. Strong at image captioning and visual question answering.
Try NowDiffusionGemma 26B
Google's novel diffusion-based language model. Uses a unique generation approach distinct from autoregressive transformers, enabling rich thinking.
Try NowGemma 4 31B
Google's dense 31-billion-parameter Gemma model with extended thinking capabilities. Excels at nuanced language tasks and detailed explanations.
Try NowGLM 5.1
ZhipuAI's powerful multilingual model with strong Chinese and English capabilities. Versatile for creative writing, translation, and conversation.
Try NowMistral Medium 3.5
Mistral's high-quality 128k-context model with high reasoning effort. Excellent for document analysis, summarization, and structured output.
Try NowMinistral 14B
Mistral's compact and fast 14B instruction model. Great balance of speed and quality for everyday tasks and quick responses.
Try NowStable Diffusion 3.5
Stability AI's latest image generation model. Create stunning 1024×1024 images from text descriptions with exceptional detail and artistic quality.
Try NowNV-Embed v1
NVIDIA's generalist text embedding model, mapping text into dense vector representations for semantic search and retrieval.
Try NowRequest Access
Want your own access to Playground.portch.no? Enter your email address and we will reach out with credentials within a few hours.