DeepSeek-V3.2-Think

DeepSeek-V3.2-Think (DeepSeek) — 128K context, Reasoning tier. Input price: $0.55/M · Output price: $2.19/M · Avg latency 48ms. Routed via Routara OpenAI-compatible endpoint with multi-region failover and metered billing.

Input price: $0.55/M · Output price: $2.19/M · TTFT: 48ms

Specifications

  • Developer: DeepSeek
  • Category: Reasoning
  • Context window: 128K
  • Input price: $0.55 / 1M
  • Output price: $2.19 / 1M
  • Avg latency: 48 ms
  • SLA grade: S+

Typical use cases

  • Multi-step reasoning and agent workflows
  • RAG pipelines with tool calling
  • Code generation, review, and test synthesis

FAQ

  • How is DeepSeek-V3.2-Think priced on Routara? — Input tokens are billed at $0.55/M and output at $2.19/M. Charges are metered per request with no separate platform fee.
  • Is DeepSeek-V3.2-Think OpenAI-compatible? — Yes. Point any OpenAI SDK to https://api.routara.ai/v1 and set model to the Routara slug shown below.
  • Does DeepSeek-V3.2-Think support streaming? — Streaming is supported when the upstream route is live. Use stream: true in chat.completions.

Related models

  • audio1.0 (/detail/audio1-0)
  • gpt-5.3-thinking-all (/detail/gpt-5-3-thinking-all)
  • gpt-5.4-thinking-all (/detail/gpt-5-4-thinking-all)

Quick integration

  • curl https://api.routara.ai/v1/chat/completions \ -H "Authorization: Bearer YOUR_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"deepseek/deepseek-v3.2-think","messages":[{"role":"user","content":"Hello"}],"stream":true}'