deepseek/deepseek-v4

Sparse-attention flagship for coding and agents.

TextProgrammingAgents & toolsLong contextcheapopen-sourcetools

Specifications

Context window: 1.049M
Max output tokens: 65,536
Released: 2026-01
Infra providers: 6
Throughput: 132 tok/s
Latency (TTFB, sample): 420 ms
Capabilities: Supports tools

Pricing

per million tokens

Input

$0.35

Output

$1.40

Blended (3:1)

$0.61

Use this model

Works with any OpenAI SDK — same request shape, different model id.

curl https://api.lokarouter.id/v1/chat/completions \  -H "Authorization: Bearer sk-lkr-your-key" \  -H "Content-Type: application/json" \  -d '{    "model": "deepseek/deepseek-v4",    "messages": [{ "role": "user", "content": "Hello" }]  }'

Sample data for demonstration — connect your workspace for live stats.