DeepSeek-V4.1-Flash (1017)
Aggregated modalities
Channel
Each channel is compared at its lowest input-price tier; output breaks ties. A dash means /api/pricing does not provide that rate. Selecting a channel updates the endpoints and call example below.
ChannelUSD per 1M tokensInputOutputCache readCache write
deepseek/deepseek-v4.1-flash-1017Input
$0.30 USD per 1M tokens
Output
$1.20 USD per 1M tokens
Cache read
$0.006 USD per 1M tokens
Cache write
- USD per 1M tokens
Endpoints
POST
/v1/chat/completionsChat CompletionsAPI call spec
Observed from real test calls on this channel. Names, paths and values are shown exactly as the API uses them. Verified by test calls on Oct 19, 2026.
POST/v1/chat/completions
Streaming supportedUsage reported in the streamStream optionsinclude_usage
InputSystem message: YesImage as data URL: Yes
Request parameters 24 supported parameters
Core
messagesarraybodyRequiredSupportedEffect observedmessages[].role=systemarraybodySupportedEffect observedmodelstringbodyRequiredSupportedEffect observednintegerbodyNot supported
Output length
max_tokensintegerbodySupportedEffect observedAccepted values384000384001Rejected values1536000- Maximum
- 384001
max_completion_tokensintegerbodyNot supported
Sampling
frequency_penaltynumberbodySupportedValidatedAccepted values-202Rejected values-33- Minimum
- -2
- Maximum
- 2
logprobsbooleanbodySupportedEffect observedpresence_penaltynumberbodySupportedValidatedAccepted values-202Rejected values-33- Minimum
- -2
- Maximum
- 2
seedintegerbodySupportedAcceptedAccepted values1temperaturenumberbodySupportedValidatedAccepted values00.712Rejected values-13- Minimum
- 0
- Maximum
- 2
top_kintegerbodySupportedAcceptedAccepted values120top_logprobsintegerbodySupportedEffect observedtop_pnumberbodySupportedValidatedAccepted values0.10.51Rejected values-0.52- Minimum
- 0.1
- Maximum
- 1
Stop sequences
stoparraybodySupportedEffect observed
Reasoning
reasoning_effortstringbodySupportedValidatedAccepted valueshighlowmediumRejected valuesbogusthinking.typestringbodySupportedEffect observedAccepted valuesdisabledenabledRejected valuesbogus- Switches reasoning on or off
Structured output
response_format.type=json_objectstringbodySupportedEffect observedresponse_format.type=json_schemastringbodyNot supported
Tools
parallel_tool_callsbooleanbodySupportedAcceptedtool_choiceobjectbodySupportedEffect observedAccepted valuesautononeRejected values{"type":"function","function":{"name":"get_weather"}}requiredtoolsarraybodySupportedEffect observedtools[].type=functionarraybodySupportedEffect observed
Streaming
streambooleanbodySupportedEffect observedstream_options.include_usagebooleanbodySupportedEffect observed
Multimodal input
messages[].content[].type=image_urlarraybodySupportedEffect observed
Metadata
userstringbodySupportedAcceptedAccepted valuesops-probe
Response fields 49 fields
choicesarrayAlways returnedStreamed in each chunkchoices[].delta.contentstringConditional · when streamingStreamed in each chunkchoices[].delta.reasoning_contentstringConditional · when streamingStreamed in each chunkchoices[].delta.rolestringConditional · when streamingStreamed in each chunkchoices[].finish_reasonstringAlways returnedStreamed in each chunkchoices[].indexintegerAlways returnedStreamed in each chunkchoices[].logprobsnullAlways returnedStreamed in each chunkchoices[].logprobs.contentarrayConditional · when log probabilities are requestedchoices[].logprobs.content[].bytesarrayConditional · when log probabilities are requestedchoices[].logprobs.content[].bytes[]integerConditional · when log probabilities are requestedchoices[].logprobs.content[].logprobnumberConditional · when log probabilities are requestedchoices[].logprobs.content[].tokenstringConditional · when log probabilities are requestedchoices[].logprobs.content[].top_logprobsarrayConditional · when log probabilities are requestedchoices[].logprobs.content[].top_logprobs[].bytesarrayConditional · when log probabilities are requestedchoices[].logprobs.content[].top_logprobs[].bytes[]integerConditional · when log probabilities are requestedchoices[].logprobs.content[].top_logprobs[].logprobnumberConditional · when log probabilities are requestedchoices[].logprobs.content[].top_logprobs[].tokenstringConditional · when log probabilities are requestedchoices[].logprobs.reasoning_contentarrayConditional · when log probabilities are requestedchoices[].logprobs.reasoning_content[].bytesarrayConditional · when log probabilities are requestedchoices[].logprobs.reasoning_content[].bytes[]integerConditional · when log probabilities are requestedchoices[].logprobs.reasoning_content[].logprobnumberConditional · when log probabilities are requestedchoices[].logprobs.reasoning_content[].tokenstringConditional · when log probabilities are requestedchoices[].logprobs.reasoning_content[].top_logprobsarrayConditional · when log probabilities are requestedchoices[].logprobs.reasoning_content[].top_logprobs[].bytesarrayConditional · when log probabilities are requestedchoices[].logprobs.reasoning_content[].top_logprobs[].bytes[]integerConditional · when log probabilities are requestedchoices[].logprobs.reasoning_content[].top_logprobs[].logprobnumberConditional · when log probabilities are requestedchoices[].logprobs.reasoning_content[].top_logprobs[].tokenstringConditional · when log probabilities are requestedchoices[].message.contentstringAlways returnedchoices[].message.reasoning_contentstringAlways returnedchoices[].message.rolestringAlways returnedchoices[].message.tool_callsarrayConditional · when tools are usedchoices[].message.tool_calls[].function.argumentsstringConditional · when tools are usedchoices[].message.tool_calls[].function.namestringConditional · when tools are usedchoices[].message.tool_calls[].idstringConditional · when tools are usedchoices[].message.tool_calls[].indexintegerConditional · when tools are usedchoices[].message.tool_calls[].typestringConditional · when tools are usedcreatedintegerAlways returnedStreamed in each chunkidstringAlways returnedStreamed in each chunkmodelstringAlways returnedStreamed in each chunkobjectstringAlways returnedStreamed in each chunksystem_fingerprintstringAlways returnedStreamed in each chunkusagenullConditional · when streamingSent in the final chunkusage.completion_tokensintegerAlways returnedSent in the final chunkusage.completion_tokens_details.reasoning_tokensintegerAlways returnedSent in the final chunkusage.prompt_cache_hit_tokensintegerAlways returnedSent in the final chunkusage.prompt_cache_miss_tokensintegerAlways returnedSent in the final chunkusage.prompt_tokensintegerAlways returnedSent in the final chunkusage.prompt_tokens_details.cached_tokensintegerAlways returnedSent in the final chunkusage.total_tokensintegerAlways returnedSent in the final chunk
Finish reasonslengthstoptool_calls
Caching
Automatic prompt caching
Cache lifetimes5m
Smallest cached prompt observed: 4,877 tokens
Usage fieldsprompt_cache_hit_tokensprompt_cache_miss_tokensprompt_tokens_details.cached_tokens
Call it
Using deepseek/deepseek-v4.1-flash-1017
1from openai import OpenAI23client = OpenAI(4 base_url="https://www.realrelay.ai/v1",5 api_key="sk-***",6)78response = client.chat.completions.create(9 model="deepseek/deepseek-v4.1-flash-1017",10 messages=[{"role": "user", "content": "Hello"}],11)1213print(response.choices[0].message.content)1import OpenAI from "openai";23const client = new OpenAI({4 baseURL: "https://www.realrelay.ai/v1",5 apiKey: process.env.REALRELAY_API_KEY,6});78const response = await client.chat.completions.create({9 model: "deepseek/deepseek-v4.1-flash-1017",10 messages: [{ role: "user", content: "Hello" }],11});1213console.log(response.choices[0].message.content);1curl https://www.realrelay.ai/v1/chat/completions \2 -H "Authorization: Bearer $REALRELAY_API_KEY" \3 -H "Content-Type: application/json" \4 -d '{5 "model": "deepseek/deepseek-v4.1-flash-1017",6 "messages": [{ "role": "user", "content": "Hello" }]7 }'Prices and availability as published on Oct 19, 2026.
Capabilities
ReasoningTool useVision
Input
TextImage
Output
Text
Frequently asked questions
- What is DeepSeek-V4.1-Flash (1017)?
- DeepSeek-V4.1-Flash (1017) is a model from DeepSeek. RealRelay.ai serves it through one OpenAI-compatible API.
- How much does DeepSeek-V4.1-Flash (1017) cost?
- DeepSeek-V4.1-Flash (1017) costs from $0.30 per 1M input tokens and $1.20 per 1M output tokens. Pricing is tiered, and this is the starting tier.
- What inputs and outputs does DeepSeek-V4.1-Flash (1017) support?
- Input: Text and Image. Output: Text.
- What is the context window of DeepSeek-V4.1-Flash (1017)?
- DeepSeek-V4.1-Flash (1017) has a context window of 1M tokens. It returns up to 384K output tokens.
- What can DeepSeek-V4.1-Flash (1017) do?
- Listed capabilities: Reasoning, Tool use, and Vision.
- How do I call DeepSeek-V4.1-Flash (1017) through the API?
- Use the model ID deepseek/deepseek-v4.1-flash-1017 with a RealRelay.ai API key on the OpenAI-compatible API.
More from DeepSeek
All 6 DeepSeek modelsDeepSeek-V4.1-Flash
- Context
- 1M
- In / 1M
- $0.2857
- Out / 1M
- $1.1429
DeepSeek-V4-Pro
- Context
- 1M
- In / 1M
- $1.2686
- Official price
$1.32 - Out / 1M
- $2.5372
- Official price
$3.96
DeepSeek-V4-Flash (0731)
- Context
- 1M
- In / 1M
- $0.4286
- Out / 1M
- $1.2857
DeepSeek-V4-Pro (0813)
- Context
- 1.05M
- In / 1M
- $1.32
- Out / 1M
- $3.96
DeepSeek-V4-Flash
Save 58%- Context
- 1M
- In / 1M
- $0.1247
- Official price
$0.30 - Out / 1M
- $0.2493
- Official price
$1.20