Gemini 4 Argon

Gemini 4 Argon is Google's frontier model that excels at deep work in complex, real-world tasks across domains including coding, finance, legal, cybersecurity, and science.

4 Argon differs from earlier Gemini models in the following ways:

  • Unified input and output token limit and long decodes: 4 Argon has a unified 1,048,576-token limit that includes both input and output tokens (where output includes thinking, function calls, and the final response) across all supported API surfaces (the GenerateContent API, the Interactions API, and the OpenAI compatibility Chat Completions API). A single synchronous request generates up to 65,536 output tokens (or 32,768 tokens for non-streaming direct HTTP requests). On the GenerateContent API (v1beta1), you can extend output up to 1,048,576 tokens (minus your input tokens) using long decodes: the Google Gen AI SDK and Agent Studio handle continuation_token automatically across requests, or you can pass continuation_token in a custom continuation loop (known issue: long decoding on GenerateContent is supported only in the v1beta1 API, which the Google Gen AI SDK uses by default). For details, see the developer's guide to Gemini 4 Argon.
  • Thinking levels: Supports thinking levels LOW, MEDIUM (default), and HIGH. Setting thinking_budget or thinking_level="MINIMAL" returns an HTTP 400 validation error.
  • Sampling parameters: Custom values for sampling parameters (temperature, top_p, and top_k) are ignored.
  • Context caching: 4 Argon supports implicit caching only (6,144-token minimum), with durable caching enabled by default. Explicit context caching isn't supported.

Try in Agent Studio Deploy example app Developer guide Pricing

Note: "Deploy example app" requires a Google Cloud project with billing and Agent Platform API enabled.
Model ID gemini-4-argon
Modalities
Text Input and output
Image Input only
Audio Input only
Video Input only
Token limits Context window 1,048,576
Maximum output tokens 1,048,576
Capabilities
Tools
APIs
Consumption options
Technical specifications Image
  • Maximum images per prompt: 3,000 (up to 100 URIs)
  • Maximum file size per file for inline data or direct uploads through the console: 7 MB
  • Maximum file size per file from Google Cloud Storage: 30 MB
  • Supported MIME types:
    image/png, image/jpeg, image/webp, image/heic, image/heif
Text
  • Maximum number of files per prompt: 3,000 (up to 100 URIs)
  • Maximum number of pages per file: 3,000
  • Maximum file size per file for the API or Cloud Storage imports: 50 MB(application/pdf) or 7 MB(text/plain)
  • Maximum file size per file for direct uploads through the console: 7 MB
  • Supported MIME types:
    application/pdf, text/plain
Video
  • Maximum video length (with audio): Approximately 45 minutes
  • Maximum video length (without audio): Approximately 1 hour
  • Maximum number of videos per prompt: 10
  • Supported MIME types:
    video/x-flv, video/quicktime, video/mpeg, video/mpegs, video/mpg, video/mp4, video/webm, video/wmv, video/3gpp
Audio
  • Maximum audio length per prompt: Approximately 8.4 hours, or up to 1 million tokens
  • Maximum number of audio files per prompt: 1
  • Speech understanding for: Audio summarization, transcription, and translation
  • Supported MIME types:
    audio/x-aac, audio/flac, audio/mp3, audio/m4a, audio/mpeg, audio/mpga, audio/mp4, audio/ogg, audio/pcm, audio/wav, audio/webm
Supported regions

Model availability

  • Global: global
  • Multi-region: us, eu

ML processing

  • Multi-region: us, eu

Provisioned Throughput

  • Global: global
  • Multi-region: us, eu

Standard PayGo

  • Global: global
  • Multi-region: us, eu

Priority PayGo

  • Global: global
  • Multi-region: us, eu
Versions
  • gemini-4-argon
    • Launch stage: GA
    • Release date: October 13, 2026
    • Retirement date†: October 13, 2027 or later
Security controls Online prediction
  • Data residency
  • CMEK
  • VPC-SC
  • AXT
Batch inference
  • Data residency
  • CMEK
  • VPC-SC
  • AXT
Context caching
  • Data residency
  • CMEK
  • VPC-SC
  • AXT
See Security controls for more information.

† Listed retirement dates refer to retirement of support in Gemini Enterprise Agent Platform. Models may remain accessible through the Gemini API after these dates have passed. The Gemini API is not a Google Cloud offering and is subject to its own terms of service. For details, see the Gemini API documentation.

Gemini 4 Argon Cyber Permissive

Gemini 4 Argon Cyber Permissive is a version of Gemini 4 Argon for defensive cybersecurity use cases, such as vulnerability discovery, reverse engineering, exploit verification, and patch generation. Gemini 4 Argon Cyber Permissive is a Google Frontier Cyber Model subject to Section 31 of the Service Specific Terms.

Gemini 4 Argon Cyber Permissive is available behind an allowlist. It is generally available when accessed directly on the Gemini Enterprise Agent Platform, but may also be used with CodeMender, which is subject to the "Pre-GA Offerings Terms" in the General Service Terms section of the Service Specific Terms, and the Additional Terms for Generative AI Preview Products. To request access, contact your Google team or apply through the Fairwind Program. Access to Gemini 3.8 Flash Cyber doesn't carry over; you must apply separately.

The following table shows how Gemini 4 Argon Cyber Permissive differs from Gemini 4 Argon:

4 Argon 4 Argon Cyber Permissive
Model ID gemini-4-argon gemini-4-argon-cyber-permissive
Access Publicly available Allowlist only (Fairwind Program)
Default thinking level MEDIUM (supports LOW, MEDIUM, HIGH) HIGH (supports LOW, MEDIUM, HIGH)
Computer use Supported Not supported
Supported regions global, us, eu global, us