← Back to API models

Gemini 3.8 Flash

by Google DeepMind · AI Pro/Ultra only · released September 2, 2026

Google's newest Flash model — $0.75/$3.75 per 1M tokens through the end of 2026, same intro rate as 3.6 and 3.7 Flash, but gated to paying AI Pro/Ultra subscribers rather than the free tier.

Input (intro) $0.75 / 1M
Output (intro) $3.75 / 1M
Context 1M tokens
Reverts $1.50/$7.50 Jan 1, 2027
Google AI Studio ↗ Updated September 7, 2026
§ API pricing

Per-token rates, at an introductory discount.

Input (intro)
$0.75/1M tokens
Prompt tokens
  • Introductory rate through Dec 31, 2026
  • Vision, audio, video billed as input
  • Cached input just $0.075 / 1M
Output (intro)
$3.75/1M tokens
Completion tokens
  • Introductory rate through Dec 31, 2026
  • Same price as siblings 3.6 & 3.7 Flash
  • Includes thinking tokens
Context
1Mtokens
Window
  • Native multimodal input
  • AI Pro / AI Ultra subscribers only
  • Not available on the free tier
Reverts Jan 1, 2027
$1.50/$7.50 per 1M
Standard rate
  • Intro pricing holds through Dec 31, 2026
  • Standard rate matches other Flash-tier models
  • No action needed — automatic on Google's end

Gemini 3.8 Flash bills $0.75 / 1M input and $3.75 / 1M output through Dec 31, 2026 as an introductory rate — identical to Gemini 3.6 and 3.7 Flash. Cached input is $0.075 / 1M. The rate reverts to a standard $1.50/$7.50 on Jan 1, 2027.

What's new in Gemini 3.8 Flash

Google released Gemini 3.8 Flash on September 2, 2026, three weeks after Gemini 3.7 Flash. It ships at the same introductory $0.75 input / $3.75 output per 1M tokens as its two immediate predecessors, Gemini 3.6 Flash and 3.7 Flash — all three now sit at an identical price point through the end of 2026. The notable change is access: 3.8 Flash is available only to Google AI Pro and AI Ultra subscribers, unlike 3.6 Flash, which remains the default model on Gemini's free tier.

As with 3.7 Flash's intro pricing, the discount is temporary: Google's docs list Jan 1, 2027 as the date rates revert to a standard $1.50/$7.50 per 1M. Build cost models accordingly if your workload extends past the new year.

Capabilities

3.8 Flash inherits the native multimodality of the Gemini family — text, image, audio, and video input across a 1M token context. Google has not published a detailed capability changelog versus 3.7 Flash at launch; treat it as an incremental refresh in the same Flash tier until more benchmark detail is published.

Typical use cases

  • AI Pro/Ultra subscribers wanting the newest Flash-tier model
  • Coding agents and IDE integrations
  • Multi-step agentic workflows and tool-use loops
  • General chat and document work at the discounted intro rate

Sibling and rival comparison

ModelInput / 1MOutput / 1MContext
Gemini 3.8 Flash (intro)$0.75$3.751M
Gemini 3.7 Flash (intro)$0.75$3.751M
Gemini 3.6 Flash (intro)$0.75$3.751M
Gemini 3.5 Flash-Lite$0.30$2.501M
GPT-5.6 Luna$0.20$1.201M

All three current-generation Flash models cost the same through 2026, so the choice mostly comes down to access: 3.8 Flash requires an AI Pro or Ultra subscription, while 3.6 Flash stays free. Flash-Lite is still the cheaper-volume option, and GPT-5.6 Luna undercuts all of them on raw price but lacks native video.

← See all Google / Gemini plans