Explore 50+ AI models for text, image, audio, code, and multimodal tasks — all in one place.
Open-weight instruction-tuned model. Excellent reasoning at small sizes, runs comfortably on consumer GPUs.
Frontier-class open model. Strong on coding, reasoning, and long-context tasks.
Industry-leading model for coding, analysis, and long-context (200K) tasks. Top of HumanEval and SWE-Bench.
Multimodal flagship — text, vision, and audio in a single model with low latency.
Compact and fast. Apache 2.0 licensed — great for production deployment on consumer hardware.
3.8B parameter model that punches well above its weight. Designed for edge and mobile deployment.
Multilingual frontier-class open model. Strong on Chinese, English, code, and math.
671B MoE model with frontier-level reasoning. Open weights, exceptional cost-to-performance ratio.
State-of-the-art text-to-image generation with photorealistic output and precise prompt adherence.
Most popular open image-generation model with massive community of fine-tunes and LoRAs.
High-quality text-to-image with strong prompt understanding. Available via OpenAI and ChatGPT.
Open-source speech-to-text model. Best-in-class transcription across 99 languages.
Fine-tuned for code generation, completion, and infilling. Strong on Python, C++, JavaScript.
Best open-source code model — competitive with GPT-4 on coding benchmarks.
Multimodal model with 2M context window — process entire codebases or hour-long videos.
More models being added every week. Know one we're missing? Submit it →