DeepSeek V3.2 API Pricing
DeepSeek · Budget · Apache 2.0 · released 2025-10. Prices in USD per 1 million tokens. Data snapshot: 2026-09-28.
The price we publish for DeepSeek V3.2 has changed 3 times since 2026-08-15, most recently on 2026-09-28 when the reference host changed (+4.8%); it is now the 68th cheapest of 142 Budget models. Buying from GMICloud rather than at the rate quoted here lowers the token price by about 26%, for a quantized fp8 build. It is rated a stable release but does not meet all four criteria of the FinOps Friendly badge.
- Input: $0.28 per 1M tokens
- Output: $0.42 per 1M tokens
- Batch API (input/output): not published
- Prompt-cache read: $0.03
- Context window: 164K
- Arena ELO: 1425 (as of 2026-09-25)
- Available on: Direct API, AWS Bedrock, OpenRouter (curated by hand, verify with the provider)
- AWS Bedrock regions: ap-northeast-1, us-east-1, us-east-2, us-west-2
DeepSeek V3.2 market position
On blended price ($0.378 per 1M tokens at 30% input, 70% output) it is cheaper than 52% of Budget models: 68th cheapest of 142. Across every category it is cheaper than 76% of priced text models in the catalogue: 68th cheapest of 281. Its Arena ELO of 1425 ranks 42nd of the 68 models that carry one, above 39% of them.
DeepSeek V3.2 licence, size and capabilities
DeepSeek V3.2 is open source: its weights are published under Apache 2.0, an OSI-approved licence with no field-of-use limit, so it can be self-hosted as well as bought by the token. Its published size is 685B parameters. Listed capabilities: Text, Reasoning, Agents. Filling its 164K context window once costs $0.05 in input tokens at list price.
Monthly cost at 100K requests, by business use case
From the after-discounts figure (prompt caching on the cached input share, batch API for async workloads, only where the provider publishes those rates), which assumes you implement both, to the list price. One figure means no discount is published.
| Use case | Tokens (in/out) | Monthly (after discounts – list) |
|---|---|---|
| Support Ticket | 1,500 / 500 | $40 – $63 |
| Knowledge Q&A | 2,000 / 800 | $54 – $90 |
| Meeting Summary | 10,000 / 1,200 | $305 – $330 |
| Marketing Content | 2,500 / 1,800 | $133 – $146 |
| Coding Task | 3,000 / 2,000 | $130 – $168 |
| Invoice Processing | 1,500 / 600 | $56 – $67 |
| Call Summary | 2,000 / 700 | $80 – $85 |
| Agent Workflow | 6,000 / 3,000 | $188 – $294 |
DeepSeek V3.2 pricing by host
A host is a company that runs the model and sells it by the token — the lab that made it, a cloud platform like Amazon Bedrock, Azure or Vertex, or an independent provider. One model, several sellers, and often several prices. This model is sold by 14 different hosts. The price quoted above is the one we publish so that models stay comparable; it is not necessarily the cheapest way to buy this model. List prices as published on 2026-09-28 via OpenRouter — negotiated rates, committed-use discounts and provisioned capacity are not shown. These hosts do not all serve the model at the same numeric precision, so the cheapest is not a like-for-like substitute for the dearest.
Cheapest host: GMICloud at $0.21 input / $0.31 output per 1M tokens, 26% below the price we publish on a blended basis (30% input, 70% output). It serves a quantized fp8 build, which is not the same product as the full-precision model. The dearest, SambaNova, charges 14.5x the cheapest.
| Host | Input $/1M | Output $/1M |
|---|---|---|
| GMICloud — fp8 | $0.21 | $0.31 |
| DeepInfra — fp4 | $0.26 | $0.38 |
| AtlasCloud — fp8 | $0.26 | $0.38 |
| Venice | $0.27 | $0.39 |
| Novita — fp8 | $0.27 | $0.40 |
| SiliconFlow — fp8 | $0.26 | $0.42 |
| Baidu (the price shown above) — fp8 | $0.28 | $0.42 |
| DigitalOcean | $0.30 | $0.96 |
| Alibaba — fp8 | $0.37 | $1.11 |
| Phala | $1.00 | $1.00 |
| Friendli | $0.50 | $1.50 |
| $0.56 | $1.68 | |
| Mara | $3.00 | $4.50 |
| SambaNova | $3.00 | $4.50 |
DeepSeek V3.2 price history
DeepSeek V3.2 has been in our daily price archive since tracking began on 2026-08-15, when it listed at $0.27 input / $0.40 output per 1M tokens. The price we publish has changed 3 times since then, 1 of them a change of reference host rather than a repricing. The latest change, on 2026-09-28, was +4.8% on the blended price (30% input, 70% output), to $0.28 input / $0.42 output (reference host changed, not a repricing). Net of all changes, the blended price is +4.8% against 2026-08-15.
| Date | Old price (in / out) | New price (in / out) | Blended change |
|---|---|---|---|
| 2026-09-28 | $0.27 / $0.40 | $0.28 / $0.42 | +4.8% (reference host changed, not a repricing) |
| 2026-08-28 | $0.26 / $0.38 | $0.27 / $0.40 | +4.9% |
| 2026-08-22 | $0.27 / $0.40 | $0.26 / $0.38 | -4.6% |
Alternatives to DeepSeek V3.2
- Next cheaper Budget model: Qwen3.8 Flash (Alibaba) at $0.374 blended, 1% less.
- Next more expensive: GPT-6 Luna (OpenAI) at $0.380 blended, 1% more.
The nearest Budget models on Arena ELO, each with its blended price and score difference against DeepSeek V3.2.
- DeepSeek V4 Flash 0423 (DeepSeek) — $0.09 in / $0.18 out per 1M, Arena ELO 1436: 60% cheaper, 11 ELO points higher
- MiniMax M3 (MiniMax) — $0.30 in / $1.20 out per 1M, Arena ELO 1440: 146% more expensive, 15 ELO points higher
- Qwen3.5-27B (Alibaba) — $0.20 in / $1.56 out per 1M, Arena ELO 1409: 204% more expensive, 16 ELO points lower
- GPT-5.6 Luna (OpenAI) — $0.20 in / $1.20 out per 1M, Arena ELO 1454: 138% more expensive, 29 ELO points higher
- Cheapest within 15 Arena points: DeepSeek V4 Flash 0423 (DeepSeek), 60% cheaper, 11 ELO points higher, out of 13 models in that band.
- Highest Arena ELO within 20% of this price: GLM 5.3 Flash (Zhipu), 4% more expensive, 49 ELO points higher, out of 3 models in that price band.
DeepSeek V3.2 head to head
Common questions about DeepSeek V3.2 pricing
- How much does DeepSeek V3.2 cost?
- DeepSeek V3.2 costs $0.28 per million input tokens and $0.42 per million output tokens on the API.
- Does DeepSeek V3.2 offer batch or cached pricing?
- Prompt-cache reads cost $0.03/M input tokens. No batch pricing is published.
- What would DeepSeek V3.2 cost per month?
- For a support-ticket workload (1,500 input + 500 output tokens per request) at 100K requests/month, DeepSeek V3.2 costs about $63 at list prices.
- Has the price of DeepSeek V3.2 changed?
- DeepSeek V3.2 has been in our daily price archive since tracking began on 2026-08-15, when it listed at $0.27 input / $0.40 output per 1M tokens. The price we publish has changed 3 times since then, 1 of them a change of reference host rather than a repricing. The latest change, on 2026-09-28, was +4.8% on the blended price (30% input, 70% output), to $0.28 input / $0.42 output (reference host changed, not a repricing). Net of all changes, the blended price is +4.8% against 2026-08-15.
These are list prices: real cost also depends on caching, batching, negotiated discounts, marketplace agreements, and existing contracts. Get a FinOps review from OptimNow or compare all models.
Build the business case for DeepSeek V3.2 in the AI ROI Calculator: add your business value on top of these costs to get ROI, break-even and payback.
All models · Compare side by side · Cloud compute pricing · Price barometer · FinOps guides · API and MCP server