HaiAI123

Curated Global AI Tools Directory

NewIndustry update
New

DeepSeek API adds web search and fetch at CNY 5 per 1,000 calls

10/18/2026 — 10/18, 22:17·2 sources (all)·2 reports (all)

Key takeaways

  • DeepSeek announced that its API now supports web search and fetch, billed per successful request at CNY 5 per 1,000 calls. Calls to the Sear…

Background & analysis

On October 17, 2026, DeepSeek announced that its Search and Fetch APIs were officially live. Both interfaces run on DeepSeek's in-house search engine and are aimed at Agent and RAG applications that need real-time web content. Search takes a query, searches the web and returns structured results, while Fetch retrieves a single page and returns its title and body as Markdown. Both are raw JSON APIs with a Base URL of api.deepseek.com.

The same day, DeepSeek swapped out the model behind its deepseek-flash API endpoint. The new checkpoint, DeepSeek-V4.1-Flash-1017, was published on Hugging Face under the MIT license, with the endpoint's name and price left unchanged. According to a DEV Community article, DeepSeek described the change as a minor update, yet all 11 benchmarks it published moved up, including a gain of 11.1 points on Terminal-Bench 4.0. The article's author said they could not reproduce DeepSWE and only checked the configuration and cache data.

On October 18, DeepSeek published a notice stating that its API now supports web search and fetching, billed by the number of successful requests at 5 yuan per 1,000 calls. The notice said that when users call the DeepSeek API from DeepSeek Harness, Claude Code or Codex, those tools already reach the Search and Fetch interfaces through server-side tool calls. Such calls are currently not billed, and per-call charges will begin at 00:00 Beijing time on November 1, 2026. Public information currently stops at this billing arrangement.

AI-generated from 2 reports · updated 1 hour ago

Latest turnAccording to a DEV Community analysis, DeepSeek replaced the model behind its deepseek-flash API on October 17 with DeepSeek-V4.1-Flash-1017, published on Hugging Face under MIT at the same name and price. DeepSeek calls it a minor update, but its benchmark table shows all 11 reported scores rising, including Terminal-Bench 4.0 up 11.1 points; the author checked the config and KV-cache math.

Related AI tools

24-hour heatHeat index 187 · peak 189 · now
24 hours agonow

Reports on this story headlines open the original

Today
  1. According to a DEV Community analysis, DeepSeek replaced the model behind its deepseek-flash API on October 17 with DeepSeek-V4.1-Flash-1017, published on Hugging Face under MIT at the same name and price. DeepSeek calls it a minor update, but its benchmark table shows all 11 reported scores rising, including Terminal-Bench 4.0 up 11.1 points; the author checked the config and KV-cache math.

    DEV Community · AIAI score 57
  2. DeepSeek said its API now supports web search and fetch, billed per successful request at 5 yuan per 1,000 calls, according to IT之家. Search and Fetch calls made through DeepSeek Harness, Claude Code or Codex are currently free and will be charged from 00:00 Beijing time on November 1, 2026. Both endpoints use DeepSeek's own search engine and return structured results or Markdown page text.

    IT之家AI score 56

Other stories people are talking about

How is heat calculated?About the method

Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.

This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.

Surge
Discussion rising fast
New
First report within 6 hours
Rising
Still gathering discussion

Back to AI industry updates →