LLM API pricing
Kimi K2.5 vs DeepSeek v3.2: price per 1M tokens
Updated 2026-10-14
Kimi K2.5 costs $0.6 / $3.00 per 1M input/output tokens (Amazon Bedrock (regional)) vs $0.62 / $1.85 for DeepSeek v3.2 (Amazon Bedrock (regional)). Blended 3:1, Kimi K2.5 is +29% vs DeepSeek v3.2.
Kimi K2.5 by channel
| Channel | Input | Output | Cache read | Paid with |
|---|---|---|---|---|
| Amazon Bedrock (regional) | $0.6 | $3.00 | n/a | AWS credits |
DeepSeek v3.2 by channel
| Channel | Input | Output | Cache read | Paid with |
|---|---|---|---|---|
| Amazon Bedrock (regional) | $0.62 | $1.85 | n/a | AWS credits |
Worked examples (monthly bill at list price)
| Workload | Kimi K2.5 | DeepSeek v3.2 | Difference |
|---|---|---|---|
| Support chatbot: 100,000 × 1,000 in / 300 out | $150.00 | $117.50 | +28% |
| RAG over docs: 50,000 × 4,000 in / 500 out | $195.00 | $170.25 | +15% |
| Coding agent: 200,000 × 20,000 in / 2,000 out | $3,600.00 | $3,220.00 | +12% |
Tokens per $25,000 of credits
- Kimi K2.5 via Amazon Bedrock (regional) (AWS credits): 20.83B tokens
- DeepSeek v3.2 via Amazon Bedrock (regional) (AWS credits): 26.95B tokens
More
Cost calculator
Monthly bill: $150
Paid with AWS credits: 166.7 months of runway.
Related cost calculators
- Amazon Bedrock AgentCore Pricing Calculator · Runtime, Gateway, Memory & Policy Costs
- Bedrock Pricing Calculator · AI Model & Tool Cost Estimator
- Amazon SageMaker AI Pricing Calculator · On-Demand Studio, Training & Inference
Prices and credit amounts are estimates and may change. Check AWS for current terms.