DeepSeek V3.1
DeepSeekReleased Aug 21, 2025
DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context...
- Context window
- 164K tokens
- Max output
- 33K tokens
- Input
- text
- Output
- text
- Tokenizer
- DeepSeek
- Knowledge cutoff
- Mar 31, 2025
- Released
- Aug 21, 2025
- Reasoning
- optional
Pricing
Per 1M tokens. The provider price and our flat 3% fee are separate columns — what you pay is their sum.
| Per 1M tokens | Provider | + 3% fee | You pay |
|---|---|---|---|
| Input | $0.210 | $0.0063 | $0.216 |
| Output | $0.790 | $0.024 | $0.814 |
| Cache read | $0.130 | $0.0039 | $0.134 |
Benchmarks
Design Arena
| Category | Elo | Win rate | Rank |
|---|---|---|---|
| 3Dmodels | 1154 | 48% | #61 |
| Codemodels | 1161 | 47.9% | #65 |
| Data vizmodels | 1143 | 46.8% | #66 |
| Game devmodels | 1150 | 47.2% | #64 |
| SVGmodels | 1024 | 38.2% | #69 |
| UI componentsmodels | 1139 | 47.5% | #64 |
| Websitesmodels | 1165 | 48% | #63 |
Head-to-head preference voting. How we filter and rank
Supported parameters
- frequency_penalty
- include_reasoning
- logit_bias
- logprobs
- max_tokens
- min_p
- presence_penalty
- reasoning
- repetition_penalty
- response_format
- seed
- stop
- structured_outputs
- temperature
- tool_choice
- tools
- top_k
- top_logprobs
- top_p
Call it
OpenAI-compatible: point your SDK at api.openkey.ai/v1 and use model deepseek/deepseek-chat-v3.1.
curl https://api.openkey.ai/v1/chat/completions \
-H "Authorization: Bearer $OPENKEY_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek/deepseek-chat-v3.1",
"messages": [{"role": "user", "content": "Hello"}]
}'import os
from openai import OpenAI
client = OpenAI(
base_url="https://api.openkey.ai/v1",
api_key=os.environ["OPENKEY_API_KEY"],
)
completion = client.chat.completions.create(
model="deepseek/deepseek-chat-v3.1",
messages=[{"role": "user", "content": "Hello"}],
)
print(completion.choices[0].message.content)import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.openkey.ai/v1",
apiKey: process.env.OPENKEY_API_KEY,
});
const completion = await client.chat.completions.create({
model: "deepseek/deepseek-chat-v3.1",
messages: [{ role: "user", content: "Hello" }],
});
console.log(completion.choices[0].message.content);Questions
- How much does DeepSeek V3.1 cost via API?
- Through OpenKey, DeepSeek V3.1 costs $0.216 per 1M input tokens and $0.814 per 1M output tokens. That is the provider price ($0.210 / $0.790) plus a flat 3% fee — nothing else.
- What is DeepSeek V3.1's context window?
- DeepSeek V3.1 accepts up to 164K tokens of context and returns up to 33K tokens per request.
- Is DeepSeek V3.1 OpenAI-compatible?
- Yes. Send requests to OpenKey's /v1/chat/completions endpoint with model "deepseek/deepseek-chat-v3.1" using any OpenAI SDK — only the base URL and API key change.
- What inputs does DeepSeek V3.1 support?
- DeepSeek V3.1 accepts text input and produces text output. Its knowledge cutoff is Mar 31, 2025.
More from DeepSeek
All DeepSeek models →DeepSeek V4 Flash
DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window. It is designed for fast inference and...
ctx 1.0Min $0.093out $0.185
DeepSeek V4 Pro
DeepSeek V4 Pro is a large-scale Mixture-of-Experts model from DeepSeek with 1.6T total parameters and 49B activated parameters, supporting a 1M-token context window. It is designed for advanced reasoning, coding,...
ctx 1.0Min $0.448out $0.896
DeepSeek V3.2
DeepSeek-V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism...
ctx 131Kin $0.236out $0.353
DeepSeek V3.2 Exp
DeepSeek-V3.2-Exp is an experimental large language model released by DeepSeek as an intermediate step between V3.1 and future architectures. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism...
ctx 164Kin $0.278out $0.422