OptimTokenOptimToken › Models by maker › DeepSeek
DeepSeek API pricing
As of 2026-10-05, the OptimToken catalogue holds 14 DeepSeek models. By price tier, 3 are Mid-tier and 11 Budget. The cheapest is DeepSeek V4 Pro 0423 at $0.21 per 1M input tokens and $0.42 per 1M output tokens. The most expensive is DeepSeek V4 Pro 0813 at $0.45 input and $5.00 output, 10x the cheapest on blended price (30% input, 70% output). These are the prices we quote: a model sold by several hosts has other rates, listed on its own page.
DeepSeek models and prices
| Model | Input | Output | Batch (in / out) | Cache read | Context | Licence |
|---|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | $0.21 | $0.42 | not published | $0.02 | 1M | Open source (Apache 2.0) |
| DeepSeek V3.2 Exp | $0.27 | $0.41 | not published | not published | 164K | Open source (Apache 2.0) |
| DeepSeek V3.2 | $0.28 | $0.42 | not published | $0.03 | 164K | Open source (Apache 2.0) |
| DeepSeek V4 Flash Vision Exp | $0.22 | $0.65 | not published | $0.0069 | 1M | Open source (Apache 2.0) |
| DeepSeek V3.1 | $0.25 | $0.95 | not published | $0.13 | 164K | Open source (Apache 2.0) |
| DeepSeek V3.1 Terminus | $0.27 | $1.00 | not published | not published | 164K | Open source (Apache 2.0) |
| DeepSeek V3 | $0.26 | $1.03 | not published | not published | 164K | Open source (Apache 2.0) |
| DeepSeek V3 0324 | $0.29 | $1.14 | not published | $0.11 | 164K | Open source (Apache 2.0) |
| DeepSeek V4 Flash 0731 | $0.02 | $1.28 | not published | $0.02 | 1M | Open source (Apache 2.0) |
| DeepSeek V4 Flash 0423 | $0.03 | $1.28 | not published | $0.03 | 1M | Open source (Apache 2.0) |
| DeepSeek V4.1 Flash | $0.30 | $1.20 | $0.11 / $0.34 | $0.0060 | 1M | Open source (Apache 2.0) |
| R1 0528 | $0.50 | $2.15 | not published | $0.35 | 164K | Open source (Apache 2.0) |
| R1 | $0.70 | $2.50 | not published | not published | 64K | Open source (Apache 2.0) |
| DeepSeek V4 Pro 0813 | $0.45 | $5.00 | not published | $0.40 | 1M | Open source (Apache 2.0) |
Published discounts: a prompt-cache read rate for 10 of the 14 and a batch rate for 1 of the 14. The most recent by release date is DeepSeek V4.1 Flash (2026-09). Arena ELO scores (as of 2026-10-02) exist for 4 of the 14; the highest is DeepSeek V4 Pro 0423 at 1458.
Licences
All 14 are open source: their weights are published under the Apache 2.0 licence, OSI-approved with no field-of-use limit, so they can be self-hosted as well as bought by the token. Licences are as recorded in this catalogue: check the maker's own terms before relying on one.
Price changes since tracking began
Our daily price archive began on 2026-08-15 and holds a record for all 14 models. The price we quote has changed for 11 of them, 145 times in all, 41 of those changes flagged as a change of quoted host rather than a repricing. The most recent changes were on 2026-10-05, to 4 models. Since tracking began, 3 of them joined the catalogue, most recently DeepSeek V4.1 Flash on 2026-09-11. Last checked against the 2026-10-05 snapshot.
Each row is a model's most recent change. Its own page has the full history.
| Model | Date | Old price (in / out) | New price (in / out) | Blended change | Changes in all |
|---|---|---|---|---|---|
| DeepSeek V3 0324 | 2026-10-05 | $0.25 / $1.00 | $0.29 / $1.14 | +14.2% (quoted host changed, not a repricing) | 6 |
| DeepSeek V4 Flash 0423 | 2026-10-05 | $0.02 / $1.28 | $0.03 / $1.28 | +0.3% | 38 |
| DeepSeek V4 Pro 0813 | 2026-10-05 | $0.55 / $4.20 | $0.45 / $5.00 | +17.1% | 24 |
| DeepSeek V4.1 Flash | 2026-10-05 | $0.0030 / $2.40 | $0.30 / $1.20 | -44.7% (quoted host changed, not a repricing) | 13 |
| DeepSeek V4 Flash 0731 | 2026-10-04 | $0.02 / $1.28 | $0.02 / $1.28 | -0.1% | 22 |
| DeepSeek V3.1 Terminus | 2026-10-03 | $0.30 / $1.00 | $0.27 / $1.00 | -1.1% (quoted host changed, not a repricing) | 3 |
| DeepSeek V4 Pro 0423 | 2026-10-02 | $0.78 / $1.57 | $0.21 / $0.42 | -73.3% | 23 |
| DeepSeek V4 Flash Vision Exp | 2026-09-29 | $0.44 / $1.32 | $0.22 / $0.65 | -51.0% (quoted host changed, not a repricing) | 5 |
Another 3 models changed price earlier.
Head-to-head comparisons
Models from DeepSeek appear on 5 ready-made side-by-side pages.
- Llama 4 Maverick vs DeepSeek V3.2
- DeepSeek V4 Flash 0731 vs Gemini 3.7 Flash
- GLM 5.2 vs DeepSeek V4 Flash 0731
- DeepSeek V4 Pro 0813 vs Qwen3.7 Max
- Claude Sonnet 5 vs DeepSeek V4 Pro 0813
Other makers
- AionLabs API pricing
- Alibaba (Qwen) API pricing
- Amazon (Nova) API pricing
- Anthropic (Claude) API pricing
- ByteDance (Seed) API pricing
- Cohere (Command) API pricing
- Google (Gemini) API pricing
- inclusionAI (Ling) API pricing
- Meta API pricing
- MiniMax API pricing
- Mistral API pricing
- Moonshot (Kimi) API pricing
- Nvidia (Nemotron) API pricing
- OpenAI (GPT) API pricing
- Perplexity (Sonar) API pricing
- Sakana (Fugu) API pricing
- Tencent (Hy) API pricing
- Upstage (Solar) API pricing
- xAI (Grok) API pricing
- Xiaomi (MiMo) API pricing
- Zhipu (GLM) API pricing
Models by maker: every model in the catalogue, including makers with fewer than 3.
Your real cost also depends on caching and batching, negotiated discounts, marketplace agreements, and the contracts you already have. OptimNow audits exactly that.
Get a FinOps review