DeepSeek V3.1 API pricing
Updated
Among the 7 providers we checked (prices read 2026-11-14 to 2026-11-15), the lowest blended standard price for DeepSeek V3.1 is DeepInfra at USD 0.25 per 1M input tokens and USD 0.95 per 1M output tokens. The median across 7 standard prices is USD 0.285 / USD 1. Each price is read from the provider's own pricing page.
Does DeepSeek sell DeepSeek V3.1 directly?
DeepSeek publishes no price for DeepSeek V3.1 on its own API that we could find, so this page compares hosts only.
DeepSeek V3.1 model facts
Facts from DeepSeek's model card for deepseek-ai/DeepSeek-V3.1 and the Hugging Face model API, read 2026-11-16. BroadNet did not verify them, and hosts may serve the model at a different precision. BroadRouter is BroadNet's site.
| Fact | As stated | Source |
|---|---|---|
| Open weights | Yes. The weights are in a public Hugging Face repository (deepseek-ai/DeepSeek-V3.1) with no access gate. | Hugging Face model API, checked 2026-11-16Quote“"private": false, "gated": false, 163 .safetensors files listed” |
| Licence | MIT (mit) | DeepSeek model card metadata, checked 2026-11-16Quote“license: mit” |
| Total parameters | 671B | DeepSeek model card, checked 2026-11-16Quote“DeepSeek-V3.1 | 671B | 37B | 128K” |
| Active parameters per token | 37B | DeepSeek model card, checked 2026-11-16Quote“DeepSeek-V3.1 | 671B | 37B | 128K” |
| Context length | 128K | DeepSeek model card, checked 2026-11-16Quote“DeepSeek-V3.1 | 671B | 37B | 128K” |
| Tool calling | Yes, in non-thinking mode | DeepSeek model card, checked 2026-11-16Quote“Toolcall is supported in non-thinking mode.” |
| First published on Hugging Face (not necessarily the release date) | 2025-08-21 | Hugging Face model API, checked 2026-11-16Quote“"createdAt": "2025-08-21T02:37:52.000Z"” |
DeepSeek V3.1 price per provider
USD per 1M tokens. Standard prices are ranked on a blend of 3 input tokens to 1 output token, cheapest first; the dearest standard price is 5.1× the cheapest. Free rows, the maker's serving version and other service tiers follow, tagged and unranked.
| # | Provider | Input | Cached input | Output | Blended 3:1 | Context | Precision | Source |
|---|---|---|---|---|---|---|---|---|
| 1 | DeepInfra cheapestdeepseek-ai/DeepSeek-V3.1Notes and source quote“$0.25 in $0.95 out $0.13 cached / 1M tokens” | USD 0.25 | USD 0.13 | USD 0.95 | USD 0.425 | 163,840 | FP4 | DeepInfra page checked 2026-11-14 re-checked 2026-11-16 |
| 2 | GMI Clouddeepseek-ai/DeepSeek-V3.1Notes and source quoteNo cache-read price published. “Price (per 1M tokens) Input $0.27 Output $1.00” | USD 0.27 | not published | USD 1 | USD 0.4525 | 163,840 | FP8 | GMI Cloud page checked 2026-11-15 re-checked 2026-11-16 |
| 3 | SiliconFlowNotes and source quoteContext shown as 164K. “DeepSeek-V3.1 164K $0.27 $1.0” | USD 0.27 | not published | USD 1 | USD 0.4525 | 163,840 | not stated | SiliconFlow page checked 2026-11-14 re-checked 2026-11-16 |
| 4 | Featherlessdeepseek-ai/DeepSeek-V3.1Notes and source quoteDeveloper plan (per-token billing, from 50 USD a month in credits). No cached price published. “PRICING: INPUT $0.285 / OUTPUT $1” | USD 0.285 | not published | USD 1 | USD 0.4637 | 131,072 | FP8 | Featherless page checked 2026-11-15 re-checked 2026-11-16 |
| 5 | CoreWeave Forgedeepseek-ai/DeepSeek-V3.1Notes and source quote“Deepseek V3.1 $0.50 $1.50 $0.20” | USD 0.50 | USD 0.20 | USD 1.50 | USD 0.75 | 161,000 | not stated | CoreWeave Forge page checked 2026-11-15 re-checked 2026-11-16 |
| 6 | Google Vertex AIdeepseek-v3.1Notes and source quote“DeepSeek-V3.1 Input $0.60 … Output $1.70 … Cache Hit $0.06” | USD 0.60 | USD 0.06 | USD 1.70 | USD 0.875 | not stated | not stated | Google Vertex AI page checked 2026-11-14 re-checked 2026-11-16 |
| 7 | Microsoft AzureNotes and source quoteFoundry Models serverless (pay-as-you-go); no deployment type on the row, price shown for East US. “DeepSeek-V3.1 $1.23 N/A $4.94” | USD 1.23 | not published | USD 4.94 | USD 2.158 | not stated | not stated | Microsoft Azure page checked 2026-11-14 re-checked 2026-11-16 |
| — | Google Vertex AI batchdeepseek-v3.1Notes and source quote“DeepSeek-V3.1 Batch Input $0.30 … Batch Output $0.85” | USD 0.30 | not published | USD 0.85 | USD 0.4375 | not stated | not stated | Google Vertex AI page checked 2026-11-14 re-checked 2026-11-16 |
| — | DeepInfra flex / spotdeepseek-ai/DeepSeek-V3.1Notes and source quoteFlex service tier (0.8× standard). “Flex (0.8×) $0.20 $0.76 $0.104” | USD 0.20 | USD 0.104 | USD 0.76 | USD 0.34 | 163,840 | FP4 | DeepInfra page checked 2026-11-14 re-checked 2026-11-16 |
- Ranked: standard-tier prices only, on (3 × input + output) ÷ 4. "Cached input: not published" means the provider's page shows no cache price; budget cached tokens at the input price.
- Other tiers (batch and flex / spot): batch runs asynchronously within hours; flex and spot are best-effort and can be slow or refused at capacity; priority and fast cost more for reliability or speed. Shown, not ranked.
- Automated re-check 2026-11-16: a weekly job re-fetches each source page and looks for the recorded quote or prices. It is not a human review. "Re-checked" means it still found them. Rows with no re-check note show only the date the price was read.
- Precision is shown only where the provider states it; "not stated" is unknown, not full precision.
Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).
What does DeepSeek V3.1 cost for a real workload?
Monthly cost over 30 days at each standard price, using the presets of the LLM API cost calculator. Providers without a cached price bill cached input at the input price.
- Support chatbot: 2,000 requests a day, 1,500 input tokens (60% cached), 300 output.
- RAG over documents: 1,000 requests a day, 8,000 input tokens (30% cached), 500 output.
- Coding agent (one developer): 1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning.
| Provider | Support chatbot | RAG over documents | Coding agent (one developer) |
|---|---|---|---|
| DeepInfra | USD 33.12 | USD 65.61 | USD 351.90 |
| GMI Cloud | USD 42.30 | USD 79.80 | USD 576.00 |
| SiliconFlow | USD 42.30 | USD 79.80 | USD 576.00 |
| Featherless | USD 43.65 | USD 83.40 | USD 603.00 |
| CoreWeave Forge | USD 55.80 | USD 120.90 | USD 576.00 |
| Google Vertex AI | USD 55.44 | USD 130.62 | USD 406.80 |
| Microsoft Azure | USD 199.62 | USD 369.30 | USD 2,659 |
For the support chatbot, the cheapest price (DeepInfra, USD 33.12 a month) saves USD 10.53 against the median (USD 43.65). For the coding agent the gap is USD 224.10 a month (39% lower).
Which DeepSeek V3.1 provider is cheapest for your traffic?
DeepInfra has the lowest input price (USD 0.25) and the lowest output price (USD 0.95), so it is the cheapest standard price at every mix of input and output tokens, before cache discounts.
| Provider | No cache hits | 50% cached | 90% cached |
|---|---|---|---|
| DeepInfra | USD 0.425 cheapest | USD 0.38 cheapest | USD 0.344 cheapest |
| GMI Cloud (no cache price) | USD 0.4525 | USD 0.4525 | USD 0.4525 |
| SiliconFlow (no cache price) | USD 0.4525 | USD 0.4525 | USD 0.4525 |
| Featherless (no cache price) | USD 0.4637 | USD 0.4637 | USD 0.4637 |
| CoreWeave Forge | USD 0.75 | USD 0.6375 | USD 0.5475 |
| Google Vertex AI | USD 0.875 | USD 0.6725 | USD 0.5105 |
USD per 1M tokens, 3 input tokens to 1 output token, each provider's cheapest standard row; 1 dearer providers are in the price table above. A provider without a cached price bills cached input at the input price. Any charge for writing to the cache, which some providers add, is not included.
3 of 7 providers publish a cached-input price below the input price: DeepInfra (48% off), CoreWeave Forge (60% off), Google Vertex AI (90% off). GMI Cloud, SiliconFlow, Featherless and Microsoft Azure publish none, so cached tokens there bill at the input price.
Context windows differ: 163,840 tokens at DeepInfra, GMI Cloud and SiliconFlow; 161,000 tokens at CoreWeave Forge; 131,072 tokens at Featherless. Google Vertex AI and Microsoft Azure do not state one. The cheapest row is not always the one that fits a long prompt.
What does DeepSeek V3.1 cost in other languages?
The same text needs a different number of tokens in each language. With the DeepSeek V4.1 Flash tokenizer, sending 1M characters costs this much at the cheapest standard input price we found for DeepSeek V3.1:
| Content | Input per 1M characters | Cheapest provider |
|---|---|---|
| English | USD 0.054 | DeepInfra |
| Arabic MSA | USD 0.094 | DeepInfra |
| Hindi | USD 0.144 | DeepInfra |
| Chinese | USD 0.147 | DeepInfra |
| Code | USD 0.066 | DeepInfra |
From BroadRouter's real cost by language dataset, which covers 15 languages and content types for every model we could match to a measured tokenizer.
Estimate your own bill
- LLM API cost calculator with DeepSeek V3.1 preloaded: your traffic and cache rate, against any of 428 models.
- DeepSeek token counter: count the tokens in your real prompts. Tokenizers differ, so the same text costs a different number of tokens on different models.
- Prompt caching calculator: what a higher cache-hit rate saves.
How does DeepSeek V3.1 compare with other DeepSeek models?
The DeepSeek models whose cheapest standard price is closest to DeepSeek V3.1's (USD 0.425 blended):
| Model | Cheapest input | Cheapest output | Max context | Providers |
|---|---|---|---|---|
| DeepSeek V3 | USD 0.25 | USD 1 | 163,840 | 3 |
| DeepSeek-V3.1-Terminus | USD 0.27 | USD 1 | 163,840 | 3 |
| DeepSeek R1 Distill Llama 70B | USD 0.25 | USD 0.75 | 131,072 | 2 |
| DeepSeek V3 0324 | USD 0.29 | USD 1.14 | 131,072 | 3 |
| DeepSeek-V4-Flash-Vision-Exp | USD 0.2156 | USD 0.6468 | 1,048,576 | 3 |
| DeepSeek-V3.2-Exp | USD 0.27 | USD 0.41 | 163,840 | 2 |
Priced closest from other makers: Qwen3 Coder 480B A35B (Qwen, USD 0.4375) · Step 3.7 Flash (StepFun, USD 0.4375) · MiniMax M3 (MiniMax, USD 0.4125).
Compare DeepSeek V3.1 with
- DeepSeek V3.1 vs DeepSeek V3: API price side by side
- DeepSeek V3.2 vs DeepSeek V3.1: API price side by side
Also compared: vs DeepSeek V3 · vs DeepSeek V3.2.
Every DeepSeek V3.1 price comparison · All model-vs-model price comparisons
All 428 models: API prices compared · Cheapest LLM APIs right now · LLM API price index (open data) · Download all prices (CSV, JSON) · Price change log
Questions
How much does the DeepSeek V3.1 API cost per token?
From USD 0.25 per 1M input tokens (USD 0.00000025 per token) and USD 0.95 per 1M output tokens (USD 0.00000095 per token) at the lowest standard prices we found. The median of 7 standard prices is USD 0.285 input and USD 1 output. Each price was read from the provider's own pricing page; the source is linked in its row.
Which provider is cheapest for DeepSeek V3.1?
Among the 7 providers we checked (prices read 2026-11-14 to 2026-11-15), the lowest is DeepInfra, ranked on a blend of 3 input tokens to 1 output token (USD 0.425 per 1M). It also has the lowest input and the lowest output price. Batch, flex, priority and fast tiers are shown in the table but not ranked.
Does DeepSeek sell DeepSeek V3.1 on its own API?
DeepSeek publishes no price for DeepSeek V3.1 on its own API that we could find, so there is no maker price to compare. The lowest-priced host we checked is DeepInfra at USD 0.25 / USD 0.95 per 1M tokens.
Does DeepSeek V3.1 have cached input pricing?
3 of 7 standard prices include one. The lowest is USD 0.06 per 1M at Google Vertex AI; the median is USD 0.13. Providers without one bill cached input at the input price. What cached input means.
What is the DeepSeek V3.1 context window?
Up to 163,840 tokens. 2 standard prices come with less, down to 131,072 tokens (Featherless and CoreWeave Forge). Check the context column before you pick the cheapest row for long prompts. Context window calculator.
How much does a coding agent cost per month on DeepSeek V3.1?
For 1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning, over 30 days: USD 351.90 a month at DeepInfra, USD 576.00 at the median price and USD 2,659 at Microsoft Azure. Try your own numbers.
Is there a batch, flex or priority price for DeepSeek V3.1?
Yes: batch from Google Vertex AI at USD 0.30 / USD 0.85; flex / spot from DeepInfra at USD 0.20 / USD 0.76 per 1M input / output tokens. These tiers trade speed or availability for price (batch runs within hours, flex and spot can be slow or refused at capacity, priority and fast cost more for speed or reliability), so they are shown but not ranked.
Is DeepSeek V3.1 free to use through an API?
No provider in our data lists a free price for DeepSeek V3.1 as of 2026-11-15. Free tiers and credits by provider (checked 2026-11-14) lists general offers that do not name a model; all free LLM API models are on one page.
Is DeepSeek V3.1 open source, with open weights, and what is its licence?
DeepSeek publishes the weights on Hugging Face as deepseek-ai/DeepSeek-V3.1, in a public repository with no access gate. The licence, as the DeepSeek model card metadata states it: MIT (mit). Open weights means you can download and run the model yourself; what you may do with it, including commercial use, is set by that licence, so read it in full (checked 2026-11-16). These facts come from DeepSeek's card and the Hugging Face model API; BroadNet did not verify them, and hosts may serve the model at a different precision.
How many parameters does DeepSeek V3.1 have?
Total parameters, as the DeepSeek model card states them: 671B. Active parameters per token: 37B. (checked 2026-11-16; BroadNet did not verify these figures.)
Does DeepSeek V3.1 support tool calling?
Yes, in non-thinking mode, according to the DeepSeek model card: “Toolcall is supported in non-thinking mode.” (checked 2026-11-16). Whether the tools parameter works through an API also depends on the provider serving it, so check that provider's docs. BroadNet did not test it.
How do I call DeepSeek V3.1 with the OpenAI SDK?
BroadNet does not serve DeepSeek V3.1. Use the OpenAI-compatible base URL of a provider that lists it: create the client with OpenAI(base_url=..., api_key=...) and pass that provider's model ID. Documented base URLs for providers on this page, in alphabetical order: DeepInfra (https://api.deepinfra.com/v1/openai, model deepseek-ai/DeepSeek-V3.1). Each provider page quotes the provider's docs with a check date.
Read as MarkdownAsk ChatGPTAsk ClaudeAsk Perplexity
Sources