Head-to-Head Pricing Benchmark
Side-by-side pricing and context window comparison for production model selection.
Default Recommendation (120M input + 60M output)
DeepSeek: DeepSeek V3 is lower-cost for the default monthly workload scenario.
Adjust the workload in the calculator below to see a live recommendation for your usage.
| Metric | DeepSeek: DeepSeek V3.2 Speciale | DeepSeek: DeepSeek V3 |
|---|---|---|
| Developer | deepseek | deepseek |
| Context Window | 163,840 | 163,840 |
| Max Output Tokens | 163,840 | 163,840 |
| Input Cost / 1M Tokens | $0.4000 | $0.3200 |
| Output Cost / 1M Tokens | $1.20 | $0.8900 |
| Projected Monthly Cost | $120 | $92 |
| Vision | ❌ No | ❌ No |
| Tool Calling | ❌ No | ✅ Yes |
| Structured Output | ✅ Yes | ✅ Yes |
| Reasoning | ✅ Yes | ❌ No |
| Web Search | ❌ No | ❌ No |
| Tokenizer | DeepSeek | DeepSeek |
| MMLU Score | 86.3 | N/A |
| GPQA | 87.1 | N/A |
Price History
Current Input / 1M
$0.4000
Current Output / 1M
$1.20
Price History
Current Input / 1M
$0.3200
Current Output / 1M
$0.8900
Performance History
Current TPS
0.00
Current Latency
0ms
Uptime
100.0%
Performance History
Current TPS
0.00
Current Latency
0ms
Uptime
98.8%
Adjust your workload to see projected monthly costs.
DeepSeek: DeepSeek V3.2 Speciale
$120
per month
DeepSeek: DeepSeek V3
$92
per month
Lower costLive Recommendation
DeepSeek: DeepSeek V3 is lower-cost at 120M input + 60M output tokens/month.
Continue evaluation with more “A vs B pricing” decision pages.