DeepSeek-V3 by DeepSeek scores 22.8 out of 100 on the Neura Intelligence Index, ranking #298 among tracked AI models. This composite score is computed from 3 individual benchmarks across coding, math, reasoning, and language categories.

DeepSeek-V3 costs $0.32/1M input tokens and $0.89/1M output tokens via API. At a typical 3:1 input-to-output ratio with 1M tokens/day usage, the estimated monthly cost is approximately $14.

Yes, DeepSeek-V3 has open weights available under the mit_+_model_license_(commercial_use_allowed) license. You can download and self-host the model.

DeepSeek-V3 supports a context window of 164K tokens, allowing it to process moderate-length documents and conversations in a single request.

DeepSeek-V3 achieves its highest score on GPQA Diamond with a score of 59.1%. This benchmark evaluates 448 graduate-level science questions (biology, chemistry, physics) designed to be hard for non-experts.