Apps on VibeLeaderboard
DeepSeek's theorem-proving language model that combines RL and Monte Carlo tree search to solve formal math problems in Lean.
DeepSeek's open-source LLM specialized in formal theorem proving in Lean 4.
DeepSeek's hierarchical 3D generation model — produces detailed textured 3D assets from text prompts with multi-stage refinement.
DeepSeek's mixture-of-experts vision-language model series for advanced multimodal understanding — strong document, OCR, and chart performance.
DeepSeek's strong, economical mixture-of-experts language model — 236B params with only 21B active, competitive with GPT-4 at a fraction of the cost.
DeepSeek's mixture-of-experts code model that matches GPT-4 Turbo on code tasks — open weights, 338 programming languages, 128K context.
Open-source reasoning model from DeepSeek trained via reinforcement learning, with chain-of-thought reasoning competitive with closed frontier models.
DeepSeek's vision-language model for OCR via "contexts optical compression" — turning long-context text problems into image understanding.
DeepSeek's open-source code LLM family trained on 2T tokens with strong performance across 80+ programming languages.
A high-performance CUDA library for FP8/FP4 tensor operations in large language models, featuring optimized GEMM kernels, MoE fusion, and runtime JIT compilation. Designed for NVIDIA GPUs with clean, accessible code for learning GPU optimization techniques.
DeepEP DeepEP is a communication library tailored for Mixture-of-Experts (MoE) and expert parallelism (EP). It provides high-throughput and low-latency all-to-all GPU kernels, which are also known as MoE dispatch and combine. The library also supports low-precision operations, including FP8. To align with the group-limited gating algorithm proposed in the DeepSeek-V3 paper, DeepEP offers a set of kernels optimized for asymmetric-domain bandwidth
This appears to be an AI model/framework from DeepSeek AI company.