Models · Model provider

DeepSeek

DeepSeek develops downloadable and hosted models for reasoning, tool use and multimodal workloads.

+ Follow

Builder facts

  • Model IDs and inputs. The direct API serves V4.1-Flash-1017 as deepseek-flash and V4-Pro-0813 as deepseek-v4-pro. Both support one-million-token context, thinking and tool calls. Flash also accepts images. Source ↗
  • Direct API billing. DeepSeek rates vary by cache hit, cache miss and output tokens and by peak or off-peak period. Source ↗
  • Search and fetch billing. Standalone Search and Fetch calls are billed per request. Since November 18, 2026, model-triggered search and fetch have been billed only for model tokens. Source ↗
  • Downloadable V4 weights. DeepSeek publishes the V4 Pro weights under the MIT license for independent deployment. Source ↗

Sources checked

Publisher update ·

DeepSeek adds server-side tools to V4.1 Flash ↗

DeepSeek added search, URL reading and file inputs to V4.1 Flash across Messages, Responses and Chat Completions, and made OCR, transcription and speech synthesis available as standalone APIs.

Change timeline

0 stories

No verified releases in this timeline yet.