> ## Content Index
> Fetch the complete content index at: https://iteracompute.com/blog/llms.txt
> Use this file to discover other available public pages before exploring further.

# DeepSeek V4 Pro API price cut: over 54% below official peak rates
- URL: https://iteracompute.com/blog/deepseek-v4-pro-api-price-cut/
- Published: 2026-10-25T04:54:18.000Z
- Updated: 2026-10-26T01:19:58.000Z
- Description: Our listed Pro 0813 rates are now $0.60 input and $1.79 output per million tokens. See the reductions and compare official peak-period costs.
- Author: IteraCompute

We've lowered our listed input and output prices for **DeepSeek V4 Pro 0813**. IteraCompute's current standard rates are **$0.60 per million input tokens** and **$1.79 per million output tokens**.

Compared with DeepSeek's official peak-period prices for the same 0813 version, those rates are approximately **54.5% lower for input** and **54.8% lower for output**. For an illustrative request with 20,000 input tokens and 2,000 billable output tokens, the total is **$0.01558 on IteraCompute**, versus **$0.03432 at official peak rates**.

**Prices verified October 25, 2026.** All prices below are in USD per 1 million tokens. This article compares standard input and billable output charges only. Check the [current Pro 0813 model page](https://iteracompute.com/models/deepseek-v4-pro/) and your Console before sending traffic.

## What changed on IteraCompute?

Relative to our October 23 catalog, the listed input rate has fallen from $0.79 to $0.60 per million tokens, while output has fallen from $2.38 to $1.79.

| Token category | October 23 catalog / 1M | Current rate / 1M | Reduction |
| -------------- | ----------------------- | ----------------- | --------- |
| Input          | $0.79                   | $0.60             | 24.1%     |
| Output         | $2.38                   | $1.79             | 24.8%     |

The cost reduction for a complete request depends on its input-to-output mix. Use the two rates together when estimating a workload.

## Compare with DeepSeek's official peak prices

As checked on October 25, 2026, the [official DeepSeek price table](https://api-docs.deepseek.com/quick%5Fstart/pricing/) maps `deepseek-v4-pro` to **DeepSeek-V4-Pro-0813**. IteraCompute exposes that version as `deepseek/deepseek-v4-pro-0813`.

| DeepSeek V4 Pro 0813                         | Input / 1M | Output / 1M |
| -------------------------------------------- | ---------- | ----------- |
| IteraCompute — current listed standard rates | $0.60      | $1.79       |
| DeepSeek official — peak periods             | $1.32      | $3.96       |
| IteraCompute reduction versus official peak  | 54.5%      | 54.8%       |

Official peak periods are **01:00–04:00 and 06:00–10:00 UTC, Monday–Friday, excluding Chinese public holidays**. In Beijing time (UTC+8), those windows are 09:00–12:00 and 14:00–18:00\. The comparison applies to that official pricing tier, not every billing period.

## A worked example: 20k input + 2k output

For this illustration, every input token is billed at the listed standard input rate. Output means total billable output, including any billable reasoning output.

```
Total cost in USD =
  input tokens × input rate / 1,000,000
  + billable output tokens × output rate / 1,000,000
```

| Cost component               | IteraCompute | DeepSeek official peak |
| ---------------------------- | ------------ | ---------------------- |
| 20,000 input tokens          | $0.01200     | $0.02640               |
| 2,000 billable output tokens | $0.00358     | $0.00792               |
| **Total**                    | **$0.01558** | **$0.03432**           |

Under these assumptions, IteraCompute's total is about **54.6% lower**. At 1,000 requests with identical token counts, the illustrative totals are **$15.58 versus $34.32**, a difference of **$18.74**.

These are arithmetic examples, not measured bills. They exclude separate tool fees, promotions and account-specific adjustments. Count all calls and retries in an agent workflow, and settle against actual usage and the applicable billing terms.

## Start using Pro 0813 through IteraCompute

[Create an IteraCompute account](https://platform.iteracompute.com/sign-up), check the model's access and balance requirements, and create an API key in the Console. Keep the key on your server.

Use the OpenAI-compatible base URL `https://api.iteracompute.com/v1` and set `model` to `deepseek/deepseek-v4-pro-0813`. Follow the maintained [Pro 0813 OpenAI SDK guide](https://iteracompute.com/integrations/openai-sdk/) for Python or JavaScript, then inspect the reply and usage from a small request.

See the [API documentation](https://iteracompute.com/docs.html#api-reference) for authentication, streaming and error handling.

**Sources, checked October 25, 2026:** IteraCompute's [public Models API](https://api.iteracompute.com/v1/models) and [Pro 0813 page](https://iteracompute.com/models/deepseek-v4-pro/), and DeepSeek's [official Models & Pricing page](https://api-docs.deepseek.com/quick%5Fstart/pricing/). Earlier IteraCompute rates are from our October 23 catalog snapshot. Price comparisons do not establish identical features, service limits or response times.