Skip to main content
deepseek-v3-2 is the older V3 series of the DeepSeek family. It’s still a capable general model, but the V4 family is newer, smarter, and adds a reasoning chain.
  • ✅ Solid general-purpose DeepSeek model.
  • ❌ V3, not V4 — the V4 family is newer, smarter, and supports reasoning chain.
If you want the latest reasoning capability, go to deepseek-v4-pro. If you have prompts already tuned against V3, this id keeps them running. Pricing: see the rate card.

​
Protocols

ProtocolPath
OpenAI Chat CompletionsPOST https://llm.bytespike.ai/v1/chat/completions
Anthropic Messages is not supported on this model — V3.2 only exposes the OpenAI Chat Completions shape. For Anthropic Messages with DeepSeek, use deepseek-v4-pro.

​
Quickstart

curl https://llm.bytespike.ai/v1/chat/completions \
  -H "Authorization: Bearer $BYTESPIKE_API_KEY" \
  -H "content-type: application/json" \
  -d '{
    "model": "deepseek-v3-2",
    "messages": [
      { "role": "user", "content": "Hello, ByteSpike." }
    ]
  }'

​
Capabilities

CapabilitySupported
Chat Completions✅
Anthropic Messages— (use deepseek-v4-pro for Anthropic)
Streaming (SSE)✅
Tool use✅
JSON mode✅
Reasoning chain— (V3 series; V4 has reasoning)
Vision (HTTP API)❌
Context window64K tokens

​
When to use

  • V3-tuned prompts — if you’ve optimized prompts against V3 and don’t want to migrate to V4 yet.
  • General DeepSeek workloads — when you don’t need the V4 reasoning chain.
When not to use:
  • Latest DeepSeek capabilities — go to deepseek-v4-pro.
  • Anthropic Messages required — same.
  • Long context — 64K is the ceiling here.

​
Next