DeepSeek

Every recorded change to 3 public pages. Faded entries are wording-only changes.

1. Nov 0627 line(s) added, 35 removed (around "Your First API Call").api-pricingpricing+27 −35raw difflive page
2. Nov 050 line(s) added, 1 removed.api-pricingpricing+0 −1raw difflive page
3. Nov 05194 line(s) added, 26 removed (around "Change Log").newschangelog+194 −26raw difflive page
4. Nov 051 line(s) added, 1 removed.rate-limitlimits+1 −1raw difflive page

1. Oct 26DeepSeek Harness is now in developer preview and the DeepSeek API now supports OpenAI and Anthropic compatible formats.newschangelog+26 −194raw difflive page
2. Oct 2626 line(s) added, 194 removed (around "Your First API Call").newschangelog+26 −194raw difflive page
3. Oct 19DeepSeek V4.1 Flash has been upgraded to DeepSeek-V4.1-Flash-1017 with improved performance and reduced runtime while maintaining existing API pricing.newschangelog+17 −0raw difflive page
4. Oct 1917 line(s) added, 0 removed.newschangelog+17 −0raw difflive page
5. Oct 18Minor: Corrected the messages parameter format in the HTTP request body examples and added trailing commas to SDK code snippets.rate-limitlimits+4 −4raw difflive page
6. Oct 184 line(s) added, 4 removed.rate-limitlimits+4 −4raw difflive page
7. Oct 17Added Search and Fetch standalone APIs billed at $0.75 per 1,000 successful requests and updated DeepSeek-V4.1-Flash to version 1017.api-pricingpricing+11 −4raw difflive page
8. Oct 1711 line(s) added, 4 removed.api-pricingpricing+11 −4raw difflive page

1. Sep 22Updated off-peak rate definitions to exclude Chinese public holidays from peak hours and removed the DeepSeek V4 Pro service extension notice.api-pricingpricing+5 −6raw difflive page
2. Sep 225 line(s) added, 6 removed.api-pricingpricing+5 −6raw difflive page
3. Sep 19Updated concurrency limits to 2500 for deepseek-flash and 500 for deepseek-v4-pro, and updated code examples to use deepseek-flash.rate-limitlimits+8 −8raw difflive page
4. Sep 198 line(s) added, 8 removed.rate-limitlimits+8 −8raw difflive page
5. Sep 15Updated model names to deepseek-flash and deepseek-v4-pro, added specific pricing for off-peak and peak hours, and defined concurrency limits.api-pricingpricing+30 −26raw difflive page
6. Sep 1530 line(s) added, 26 removed (around "Models & Pricing").api-pricingpricing+30 −26raw difflive page
7. Sep 13Released DeepSeek-V4.1-Flash with native multimodal support, reduced API prices, and extended DeepSeek V4 Pro availability until September 14, 2026.newschangelog+27 −0raw difflive page
8. Sep 1327 line(s) added, 0 removed.newschangelog+27 −0raw difflive page
9. Sep 09Added deepseek-v4-flash-vision-exp to the supported models with a concurrency limit of 2500.rate-limitlimits+4 −4raw difflive page
10. Sep 094 line(s) added, 4 removed.rate-limitlimits+4 −4raw difflive page
11. Sep 06Removed the pricing table and added a first API call guide with OpenAI/Anthropic compatibility and example scripts.api-pricingpricing+26 −28raw difflive page
12. Sep 0626 line(s) added, 28 removed (around "Your First API Call").api-pricingpricing+26 −28raw difflive page

1. Aug 25DeepSeek-V4-Flash-Vision-Exp is now available on the DeepSeek API platform as an experimental multimodal vision understanding model.newschangelog+14 −0raw difflive page
2. Aug 2514 line(s) added, 0 removed.newschangelog+14 −0raw difflive page
3. Aug 24Added deepseek-v4-flash-vision-exp model with a concurrency limit of 2500 and pricing matching deepseek-v4-flash.api-pricingpricing+24 −23raw difflive page
4. Aug 2424 line(s) added, 23 removed.api-pricingpricing+24 −23raw difflive page
5. Aug 16Updated DeepSeek-V4-Pro-0813 pricing to include off-peak and peak rates, and added Responses API support for the Pro model.api-pricingpricing+23 −21raw difflive page
6. Aug 1623 line(s) added, 21 removed.api-pricingpricing+23 −21raw difflive page
7. Aug 16DeepSeek-V4-Pro is now in GA with native OpenAI Responses API support, three thinking effort levels, and peak/off-peak pricing starting August 16, 2026.newschangelog+20 −0raw difflive page
8. Aug 1620 line(s) added, 0 removed.newschangelog+20 −0raw difflive page
9. Aug 09DeepSeek plans to significantly increase overall API pricing in the near future, with specific details to be announced later.api-pricingpricing+1 −1raw difflive page
10. Aug 091 line(s) added, 1 removed.api-pricingpricing+1 −1raw difflive page
11. Aug 09The DeepSeek-V4-Flash API is now in public beta, accessible by setting the model name to deepseek-v4-flash.newschangelog+2 −2raw difflive page
12. Aug 092 line(s) added, 2 removed.newschangelog+2 −2raw difflive page
13. Aug 01DeepSeek will soon adopt a peak/off-peak pricing policy where prices will be 2x the regular prices during specified daily hours.api-pricingpricing+9 −5raw difflive page
14. Aug 019 line(s) added, 5 removed.api-pricingpricing+9 −5raw difflive page

1. Jul 31The DeepSeek-V4-Flash API is now in public beta with enhanced agent capabilities and native support for the Responses API format.newschangelog+13 −0raw difflive page
2. Jul 3113 line(s) added, 0 removed.newschangelog+13 −0raw difflive page

1. Jun 06Removed 75% discount pricing for deepseek-v4-pro and set the deprecation date for deepseek-chat and deepseek-reasoner to 2026/07/24 15:59 UTC.api-pricingpricing+5 −5raw difflive page
2. Jun 065 line(s) added, 5 removed.api-pricingpricing+5 −5raw difflive page
3. Jun 02DeepSeek API now sets concurrency limits at 500 for deepseek-v4-pro and 2500 for deepseek-v4-flash, with optional capacity expansion and user_id isolation.rate-limitlimits+34 −3raw difflive page
4. Jun 0234 line(s) added, 3 removed (around "Rate Limit & Isolation").rate-limitlimits+34 −3raw difflive page

1. May 29Added a concurrency limit of 2500 for some models and clarified that deepseek-v4-pro will be 1/4 of the original price after May 31, 2026.api-pricingpricing+2 −1raw difflive page
2. May 292 line(s) added, 1 removed.api-pricingpricing+2 −1raw difflive page

• api-pricing (pricing)
• rate-limit (limits)
• news (changelog)