DeepSeek-V4.1-Flash

GPT-6-Astra, running in OpenAI's Codex, solved 49 of 50 problems (98%) on TheoPhysEval, a new benchmark of unpublished theoretical physics research problems, at just under $116 for the full run. Anthropic's Opus 5.5 in Claude Code placed second at 92%. The authors say the agents now match specialists on concrete calculations but lack research "taste."

DeepSeek has released V4.1-Flash, a 552B-parameter multimodal Mixture-of-Experts model supporting 1M-token context and activating only 8B parameters during prefill. The model uses a new Causal Encoder-Decoder architecture and Compressed Sparse Attention 2 to cut global KV cache to 890 bytes per token, roughly a quarter of its predecessor.