LIVE
Google Introduces "Gemini 4 Carbon" as Next Step in Its Model Lineup28/10/26 · Google|DeepMind unveils Genie 4, a new generation of world model27/10/26 · Google DeepMind|Anthropic files IPO prospectus, with risk disclosures taking center stage26/10/26 · Anthropic|Robot policy trained only in simulation matches top real-world systems26/10/26|OpenAI unveils GPT-6.1 Terra, a model built for computer use26/10/26 · OpenAI|Google's AI gives evasive answers about who won the 2020 US election28/10/26 · Google|NVIDIA introduces Nemotron 3.5 Transcribe, a promptable multilingual transcription model28/10/26 · NVIDIA|OpenAI and ServiceNow deepen partnership on enterprise workflows28/10/26 · OpenAI|Failed startups' data finds a second life training AI models28/10/26|Anthropic expands AI for Science program to 20 foundations and research institutions28/10/26 · Anthropic|Anthropic shares early results on using Claude to accelerate drug development27/10/26 · Anthropic|GEPA-Curator: Filtering Failures to Better Train LLM Agents27/10/26|Google Introduces "Gemini 4 Carbon" as Next Step in Its Model Lineup28/10/26 · Google|DeepMind unveils Genie 4, a new generation of world model27/10/26 · Google DeepMind|Anthropic files IPO prospectus, with risk disclosures taking center stage26/10/26 · Anthropic|Robot policy trained only in simulation matches top real-world systems26/10/26|OpenAI unveils GPT-6.1 Terra, a model built for computer use26/10/26 · OpenAI|Google's AI gives evasive answers about who won the 2020 US election28/10/26 · Google|NVIDIA introduces Nemotron 3.5 Transcribe, a promptable multilingual transcription model28/10/26 · NVIDIA|OpenAI and ServiceNow deepen partnership on enterprise workflows28/10/26 · OpenAI|Failed startups' data finds a second life training AI models28/10/26|Anthropic expands AI for Science program to 20 foundations and research institutions28/10/26 · Anthropic|Anthropic shares early results on using Claude to accelerate drug development27/10/26 · Anthropic|GEPA-Curator: Filtering Failures to Better Train LLM Agents27/10/26|
Toolofficial publisher

url_context (Gemini)

Tool : A precise action the AI can call · requires gemini-api

Reads the content of URLs you pass in the prompt.

Google's web_fetch; same vigilance regarding injection.

Docs
Overall score
6.6 /10
Security: one caveat, see the card
Utility
7.0
Reliability
6.0
Security
One caveat
Ease
1 click · Comfortable

Security reflects where data goes, requested permissions and incidents on record at the verification date. It is not a guarantee of zero risk: check the deployment mode and the connections you enable.

What it does, concretely

When working with the Gemini API, analysing web content without manually copying entire pages is a common need. The url_context tool addresses this by allowing the model to directly fetch and read the content of web pages from URLs provided in the prompt.

In practice, the tool works in two stages: it first looks for content in an internal indexed cache for speed, then falls back to a live fetch if the page is not yet indexed. The model can process up to 20 URLs per request, each up to 34 MB, and it cites its sources through inline annotations in the response.

This tool is aimed at developers building applications that require data extraction, document comparison, multi-source summarisation, or analysis of code repositories and technical documentation. It can also be combined with Google Search to first find relevant pages and then analyse them in depth.

However, it is not suited for use cases requiring access to pages behind a login or paywall, YouTube videos, Google Workspace files, or resources on private or local networks. It is limited to the Gemini API and is not available on the Gemini Enterprise Agent Platform.

What it offers

  • ·Reading the content of public web pages from URLs provided directly in the prompt.
  • ·Processing up to 20 URLs per request, with a maximum content size of 34 MB per URL.
  • ·Two-stage retrieval: internal indexed cache first, then live fetch if the page is not yet indexed.
  • ·Inline citation annotations in the response, linking each passage back to its source URL.
  • ·Can be combined with Google Search to first find relevant pages and then analyse them in depth.
  • ·Can be combined with custom tools via function calling on Gemini 3 models.
  • ·Support for various formats: HTML, PDF, JSON, CSV, PNG and JPEG images, among others.
  • ·Automatic safety check on each URL before processing, with status returned in the response.

In practice

Price
Freemium
Commercial use
Commercial use: yes
Available on
api
Host
gemini-api
FR interface
No

Security and data

Rating
Security: one caveat, see the card
Capabilities
readswritesexecutesnetwork
Your data
Data with the vendor
Maintenance
Active

Links and repository checked on 25 October 2026 · added on 1 May 2025

How to install url_context (Gemini)

1. Obtain a Gemini API key at ai.google.dev and store it in an environment variable (GEMINI_API_KEY) rather than hard-coding it. 2. Install the Python SDK with: pip install google-genai (version 2.0.0 or higher required). 3. Import the client in your script: from google import genai, then initialise it with client = genai.Client(). 4. Add {"type": "url_context"} to the tools parameter of your call to client.interactions.create(). 5. Pass your URLs directly in the input field of the request, in free text or in a structured instruction. 6. Check the url_context_result field in the response to verify the retrieval status of each URL.
pip install google-genai

Official documentation ↗

In the same category

Whole category Feed (memory, context) →