Skip to content
LLM Pricing Calc
Pricing last verified

DeepSeek V3.2 pricing calculator

DeepSeek's 671B MoE model using multi-head latent attention (a very small KV cache). Pricing shown is DeepInfra's serverless rate.

Input
$0.26 / 1M
Cached input
$0.13 / 1M
Output
$0.38 / 1M
Context window
160K
Batch API
No
Reasoning
Yes

Official pricing ↗

Workload

Quick presets

1,000 tokens ≈ 750 English words.

Pro / Agent mode

Caching, batch, multi-turn loops, reasoning tokens

Cost comparison

Cheapest / month

$8.46

Llama 3.3 70B Instruct

Most expensive / month

$16.49

DeepSeek V3.2

Tokens / request

1.5K in

400 out (cheapest model)

Estimated cost per model
Remove
Llama 3.3 70B InstructCheapestMeta via DeepInfra · $0.10 / $0.32$0.00028$0.278$8.46$101.47
Qwen3-235B-A22B-Instruct-2507Alibaba via DeepInfra · $0.09 / $0.55$0.00036$0.355$10.80$129.58
gpt-oss-120bOpenAI via Together · $0.15 / $0.60$0.00047$0.465$14.15$169.73
DeepSeek V3.2DeepSeek via DeepInfra · $0.26 / $0.38$0.00054$0.542$16.49$197.83

Monthly = daily × 30.42. Prices are list prices for the standard context tier. Long-context surcharges, regional pricing and taxes are not included.Official pricing: DeepInfra ↗ · Together ↗

What DeepSeek V3.2 costs for common workloads

Monthly estimates using the presets above, at list price.

WorkloadPer request / sessionMonthlyYearly
Small Chatbot1K chats/day, short prompts$0.00027$8.21$98.55
RAG App5K queries/day, 4K-token retrieved context$0.00098$149.67$1,796
Coding Agent200 sessions/day × 25 turns$0.1953$1,188$14,256
Enterprise Scale250K requests/day, batched$0.00065$4,936$59,221

Official pricing ↗ (DeepSeek via DeepInfra) · API model ID deepseek-ai/DeepSeek-V3.2

DeepSeek V3.2 pricing FAQ

How much does DeepSeek V3.2 cost per 1M tokens?

DeepSeek V3.2 via DeepInfra costs $0.26 per 1M input tokens and $0.38 per 1M output tokens. Cached input is $0.13 per 1M. Prices were last verified on Nov 21, 2026.

How much does one DeepSeek V3.2 request cost?

A typical request with 1,000 input tokens and 500 output tokens costs about $0.00045. At 1,000 such requests per day that is $13.69 per month.

Does DeepSeek V3.2 support prompt caching?

Yes. Cached input tokens cost $0.13 per 1M, 50% less than regular input. Turn on Pro mode above to model your cache hit rate.

Is there a Batch API discount for DeepSeek V3.2?

No 50% batch discount is listed for DeepSeek V3.2 via DeepInfra.

What is the context window of DeepSeek V3.2?

DeepSeek V3.2 supports up to 160K tokens of context.

Is DeepSeek V3.2 cheaper than gpt-oss-120b?

For a 1,000-in / 500-out request, DeepSeek V3.2 costs $0.00045 and gpt-oss-120b costs $0.00045, so DeepSeek V3.2 is 0% cheaper. Quality, speed and caching behaviour differ, so compare on your own workload above.

More AI cost tools