DeepSeek · 2025-12-01
DeepSeek V3.2 pricing
DeepSeek-V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance.
Prices from OpenRouter’s models API, updated October 29, 2026
What DeepSeek V3.2 costs a month
| Workload | Standard |
|---|---|
| Support chatbot: 1,000 chats a day, 1,500 tokens in, 400 out | $17.64 |
| Daily summaries: 10,000 a day, 3,000 tokens in, 300 out | $290 |
| Coding agent: 400 calls a day, 60,000 tokens in (80% cached), 2,000 out | $66.53 |
30 days. Cached input billed at the cache-read price. Use your own numbers.
Cheaper models with a similar context window
| Model | Provider | Input /1M | Output /1M | Cache read /1M | Context | Chatbot, per month* |
|---|---|---|---|---|---|---|
| Ling 3.0 Flash | inclusionai | $0.021 | $0.063 | $0.0042 | 262K | $1.70 |
| GLM 5.3 Flash | Z.ai | $0.046 | $0.074 | $0.03 | 1.05M | $2.95 |
| Qwen3.7 Flash | Alibaba Qwen | $0.03 | $0.13 | $0.006 | 1M | $2.91 |
| Mercury 2.5 | inception | $0.04 | $0.15 | $0.004 | 260K | $3.60 |
| Laguna XS 2.1 free variant | poolside | $0.06 | $0.12 | $0.03 | 262K | $4.14 |
* Support chatbot: 1,000 chats a day, 1,500 tokens in, 400 out, 30 days, no caching. Prices in US dollars per 1M tokens.
Cheaper isn’t automatically as good. Glow’s GlowBench scores models on your kind of task before suggesting a swap.
Other DeepSeek models
| Model | Input /1M | Output /1M | Cache read /1M | Context | Chatbot, per month* |
|---|---|---|---|---|---|
| DeepSeek V4.1 Flash 1017 | $0.30 | $1.20 | $0.006 | 1.05M | $27.90 |
| DeepSeek V4.1 Flash | $0.045 | $0.50 | $0.006 | 1.05M | $8.03 |
| DeepSeek V4 Flash Vision Exp | $0.216 | $0.647 | $0.0069 | 1.05M | $17.46 |
| DeepSeek V4 Pro 0813 | $1.32 | $3.96 | $0.044 | 1.05M | $107 |
| DeepSeek V4 Flash 0731 | $0.119 | $0.238 | $0.024 | 1.05M | $8.21 |
| DeepSeek V4 Flash 0423 | $0.091 | $0.182 | $0.018 | 1.05M | $6.28 |
* Support chatbot: 1,000 chats a day, 1,500 tokens in, 400 out, 30 days, no caching. Prices in US dollars per 1M tokens.
Related
DeepSeek V3.2: questions
How much does DeepSeek V3.2 cost?
$0.28 per 1M input tokens and $0.42 per 1M output tokens, with cached input at $0.028. A 1,000-token prompt with a 500-token answer costs about $0.0005.
What is DeepSeek V3.2’s context window?
163,840 tokens, with up to 65,536 output tokens per response.
Is there a cheaper alternative to DeepSeek V3.2?
Ling 3.0 Flash costs $0.021 in and $0.063 out per 1M tokens with a similar context window. Whether it matches DeepSeek V3.2’s quality depends on the task: test it on your own prompts (Glow’s GlowBench does this per task).
Know what each model really costs you.
Connect OpenRouter and Glow shows spend by model, key and product, and flags the swap that saves the most at the same quality.
- Read-only keys
- Never shown again
- 2FA + Face ID
- 7-day trial