deepseek-flash

Prices, $ / 1M tokens

Loading data…

Model rankings

Loading data…

This model is not in the published task top-five lists. An absent position is not a zero quality score.

About these rankings

Positions come from popular-ai-models: reviews and surveys of Void API users in Russia. Published top-five positions are not numerical quality scores or independent benchmarks. Ranked by Void API request count, not quality. Updated every 12 hours.

Search requires web tools; file editing, design and 3D require appropriate tools and result verification. A ranking category does not imply those capabilities are available in every API method.

All rankings and methodology

Performance

Total time, s: —
Text TTFT, s: 1.996
Output / total time, tokens/s: 80.46
Input tokens: —
Output tokens: —
Of which reasoning: —

Measurement details

Source: supplied measurement table. Date, prompt, generation settings and repeat count were not provided. TTFT is time to the first text fragment. Tokens/s is all output tokens divided by total time, including reasoning, not post-first-token streaming speed. These observations do not guarantee future latency.

Connect this model

OpenAI-compatible JSON

{
  "model": "deepseek-flash",
  "messages": [
    {
      "role": "user",
      "content": "Hello"
    }
  ]
}

API documentation · Connection guides · Models and pricing