DeepSeek V3.1 API pricing
DeepSeek · Budget price tier · MIT · released 2025-08. Prices in USD per 1 million tokens. Data snapshot: 2026-11-14.
A support ticket workload of 100K requests a month costs $85.00 on DeepSeek V3.1; the next cheaper model in the Budget price tier for this workload, Qwen3.6 Flash, does it for $84.38. The price we quote for DeepSeek V3.1 has changed 4 times since 2026-08-15, most recently on 2026-09-08 when the quoted host changed (-43.9%); it is now the 109th cheapest of 159 models in the Budget price tier. None of its 3 hosts undercuts the price we quote. It is rated a stable release but does not meet all four criteria of the FinOps Friendly badge.
- Input: $0.25 per 1M tokens
- Output: $0.95 per 1M tokens
- Batch API (in/out): Not published
- Prompt-cache read: $0.13 per 1M tokens
- Context window: 164K tokens
- Arena ELO: 1420 as of 2026-11-09
Available on
AWS Bedrock regions: ap-northeast-1, us-east-2, us-west-2
Market position
On blended price ($0.74 per 1M tokens at 30% input, 70% output) it is cheaper than 32% of models in the Budget price tier: 109th cheapest of 159. Across every price tier it is cheaper than 64% of priced text models in the catalogue: 109th cheapest of 304. Its Arena ELO of 1420 ranks 78th of the 160 models that carry one, tied with one other, above 51% of them. With an Arena score of 1420, DeepSeek V3.1 clears OptimNow's Standard working quality threshold: Standard (Arena 1420+), used for meeting summary, invoice processing and call summary. Each threshold is a distance below the best ranked model, which scores 1530 on Arena scores as of 9 Nov 2026.
Licence, size and capabilities
DeepSeek V3.1 is open source: its weights are published under MIT, 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 671B parameters. Listed capabilities: Text, Reasoning, Agents. Filling its 164K context window once costs $0.041 in input tokens at list price.
Pricing by host
A host is a company that runs the model and sells it by the token: the model maker, a cloud platform like Amazon Bedrock, Azure or Vertex, or an independent host. One model, several hosts, and often several prices. This model is sold by 3 different hosts. The price we quote elsewhere on this page is marked. We quote one host's rate so that models stay comparable; it is not a recommendation to buy there.
The price we quote is already the cheapest of the 3 hosts: DeepInfra at $0.25 input / $0.95 output per 1M tokens. It serves a quantized fp4 build, which is not the same product as the full-precision model. The dearest, CoreWeave, charges 1.6x the cheapest.
| Host | Input $/1M | Output $/1M | Precision |
|---|---|---|---|
| DeepInfra (price we quote) | $0.25 | $0.95 | fp4 |
| SiliconFlow | $0.27 | $1.00 | fp8 |
| CoreWeave | $0.50 | $1.50 | fp8 |
List prices as published on 2026-11-14, 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.
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 DeepSeek 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 | $74 – $85 |
| Knowledge Q&A | 2,000 / 800 | $109 – $126 |
| Meeting Summary | 10,000 / 1,200 | $352 – $364 |
| Marketing Content | 2,500 / 1,800 | $228 – $233 |
| Coding Task | 3,000 / 2,000 | $247 – $265 |
| Invoice Processing | 1,500 / 600 | $89 – $95 |
| Call Summary | 2,000 / 700 | $114 – $117 |
| Agent Workflow | 6,000 / 3,000 | $385 – $435 |
Price history
DeepSeek V3.1 has been in our daily price archive since tracking began on 2026-08-15, when it listed at $0.25 input / $0.95 output per 1M tokens. The price we quote has changed 4 times since then, 3 of them a change of quoted host rather than a repricing. The latest change, on 2026-09-08, was -43.9% on the blended price (30% input, 70% output), to $0.25 input / $0.95 output (quoted host changed, not a repricing). Net of all changes, the blended price is back where it started on 2026-08-15.
| Date | Old price (in / out) | New price (in / out) | Blended change |
|---|---|---|---|
| 2026-09-08 | $0.55 / $1.65 | $0.25 / $0.95 | -43.9% (quoted host changed, not a repricing) |
| 2026-09-04 | $0.25 / $0.95 | $0.55 / $1.65 | +78.4% (quoted host changed, not a repricing) |
| 2026-09-02 | $0.55 / $1.65 | $0.25 / $0.95 | -43.9% (quoted host changed, not a repricing) |
| 2026-08-22 | $0.25 / $0.95 | $0.55 / $1.65 | +78.4% |
Alternatives to DeepSeek V3.1
2 of the 4 comparable models below cost less than DeepSeek V3.1: DeepSeek V3.2 Exp undercuts it by 50%, and scores 2 Arena points higher.
- Next cheaper model in the Budget price tier: MiMo-V2.5-Pro (Xiaomi) at $0.7395 blended, under 1% less.
- Next more expensive: Qwen3 Coder Flash (Alibaba) at $0.741 blended, under 1% more.
The nearest models in the Budget price tier on Arena ELO, each with its blended price and score difference against DeepSeek V3.1.
- Qwen3 VL 235B A22B Instruct (Alibaba): $0.21 in / $1.90 out per 1M, Arena ELO 1420 (88% more expensive, same Arena ELO)
- Qwen3 235B A22B Instruct 2507 (Alibaba): $0.14 in / $0.80 out per 1M, Arena ELO 1419 (19% cheaper, 1 ELO point lower)
- DeepSeek V3.2 Exp (DeepSeek): $0.27 in / $0.41 out per 1M, Arena ELO 1422 (50% cheaper, 2 ELO points higher)
- DeepSeek V3.1 Terminus (DeepSeek): $0.27 in / $1.00 out per 1M, Arena ELO 1417 (6% more expensive, 3 ELO points lower)
- Cheapest within 15 Arena points: DeepSeek V4 Flash 0423 (DeepSeek), 79% cheaper, 12 ELO points higher, out of 35 models in that band.
- Highest Arena ELO within 20% of this price: MiMo-V2.6-Pro (Xiaomi), under 1% cheaper, 72 ELO points higher, out of 19 models in that price band.
Other models from DeepSeek
The 8 nearest to DeepSeek V3.1 in price, of 13 other DeepSeek models in the catalogue. Prices per 1M tokens, input then output.
- DeepSeek V4 Flash 0731: $0.0434 / $0.30
- DeepSeek V3.2: $0.26 / $0.38
- DeepSeek V4.1 Flash 1017: $0.038 / $0.50
- DeepSeek V3.2 Exp: $0.27 / $0.41
- DeepSeek V3.1 Terminus: $0.27 / $1.00
- DeepSeek V3: $0.2574 / $1.0287
- DeepSeek V3 0324: $0.29 / $1.14
- DeepSeek V4 Flash Vision Exp: $0.44 / $1.32
All 14 DeepSeek models and prices
Common questions about DeepSeek V3.1 pricing
- How much does DeepSeek V3.1 cost?
- DeepSeek V3.1 costs $0.25 per 1M input tokens and $0.95 per 1M output tokens at DeepInfra, the host whose rate we quote. DeepInfra serves a quantized fp4 build, which is not the same product as the full-precision model. Prices as listed on 2026-11-14.
- Which host is cheapest for DeepSeek V3.1?
- The price we quote is already the cheapest of the 3 hosts: DeepInfra at $0.25 input / $0.95 output per 1M tokens. It serves a quantized fp4 build, which is not the same product as the full-precision model. The dearest, CoreWeave, charges 1.6x the cheapest. Host rates as listed on 2026-11-14.
- Does DeepSeek V3.1 offer batch or cached pricing?
- Yes, for prompt caching only. Cache reads cost $0.13 per 1M tokens, as listed on 2026-11-14. No batch pricing is published.
- What would DeepSeek V3.1 cost per month?
- With the Support Ticket profile (1,500 input and 500 output tokens per request), 100,000 requests a month cost about $85 on DeepSeek V3.1 at the list price quoted on 2026-11-14.
- Can DeepSeek V3.1 be self-hosted?
- Yes. DeepSeek V3.1 is open source: its weights are published under MIT, an OSI-approved licence with no field-of-use limit, so it can be self-hosted as well as bought by the token.
- Has the price of DeepSeek V3.1 changed?
- DeepSeek V3.1 has been in our daily price archive since tracking began on 2026-08-15, when it listed at $0.25 input / $0.95 output per 1M tokens. The price we quote has changed 4 times since then, 3 of them a change of quoted host rather than a repricing. The latest change, on 2026-09-08, was -43.9% on the blended price (30% input, 70% output), to $0.25 input / $0.95 output (quoted host changed, not a repricing). Net of all changes, the blended price is back where it started on 2026-08-15.
Its weights are published, which changes what you may run yourself: the cost gap between open weights and proprietary APIs.
These are list prices. Your real cost also depends on caching and batching, negotiated discounts, marketplace agreements, and the contracts you already have. OptimNow audits exactly that.
Build the business case · Get a FinOps review
LLM pricing · Models by maker · Cloud compute pricing · Compare models · Price barometer · FinOps guides · API and MCP server