Model Tracker

Gemini 3.8 Flash

Gemini 3.8 Flash is an AI model made by Google. Google released it on 2 September 2026. It is generally available.

Gemini 3.8 Flash is an AI model made by Google. Google released it on 2 September 2026. It is generally available. It takes text, image, video, audio and returns text. Through the API it costs $0.75 per 1M tokens for input and $3.75 per 1M tokens for output. Its knowledge cutoff is March 2026.

Released: 2 September 2026 — ai.google.dev, read 16 October 2026
Status: Generally available — ai.google.dev, read 16 October 2026
Takes: Text, Image, Video, Audio — ai.google.dev, read 16 October 2026
Returns: Text — ai.google.dev, read 16 October 2026
Open weights: Not confirmed: The maker's model page and model card do not state whether the weights are open or downloadable.
Licence: Not confirmed: The maker's model page and model card do not name a licence for the model.
Context window: Not confirmed: two readings disagree: 1048576 and 1000000
Longest answer: Not confirmed: two readings disagree: 65536 and 64000
Knowledge cutoff: March 2026 — deepmind.google, read 16 October 2026
API price, input: $0.75 per 1M tokens — ai.google.dev, read 16 October 2026
API price, output: $3.75 per 1M tokens — ai.google.dev, read 16 October 2026
Where: API, Consumer app, Cloud marketplaces — deepmind.google, read 16 October 2026
Parameters: Not confirmed: The maker does not publish a parameter count on the model page or model card.
Model card: https://deepmind.google/models/model-cards/gemini-3-8-flash/ — deepmind.google, read 16 October 2026

Benchmarks

Measured by others, not by the maker: each score with the name the measurer gives the model and the setting it ran, its day and its licence.

Arena

Text Arena, Gemini 3.8 Flash (High): 1,496 (1,492–1,501, 37,454 votes), published 15 October 2026
Code Arena (WebDev), Gemini 3.8 Flash (High): 1,586 (1,578–1,593, 10,730 votes), published 18 October 2026
Vision Arena, gemini-3.8-flash-high: 1,289 (1,279–1,300, 3,792 votes), published 9 October 2026

Arena’s leaderboard dataset (lmarena-ai/leaderboard-dataset on Hugging Face), CC BY 4.0: ratings from people’s votes between two anonymous models, as Arena publishes them. Data · CC BY 4.0

Epoch AI

Epoch Capabilities Index, Gemini 3.8 Flash: 157.2 (155.0–160.5), as of 19 October 2026
GPQA Diamond, gemini-3.8-flash_high: 95.4% (± 1.4), run 2 September 2026
FrontierMath, tiers 1–3, gemini-3.8-flash_high: 68.4% (± 2.8), run 2 September 2026
FrontierMath, tier 4, gemini-3.8-flash_high: 22.0% (± 6.5), run 2 September 2026
OTIS Mock AIME 2024–2025, gemini-3.8-flash_high: 98.9% (± 0.9), run 2 September 2026
SimpleQA Verified, gemini-3.8-flash_high: 69.7% (± 1.5), run 2 September 2026

Epoch AI, ‘Capabilities & benchmarking’, epoch.ai/benchmarks, CC BY 4.0: Epoch AI’s own runs and its Capabilities Index. Data · CC BY 4.0

Questions

When was Gemini 3.8 Flash released?

Google released Gemini 3.8 Flash on 2 September 2026.

Is Gemini 3.8 Flash still available?

Yes: it is generally available, as Google states it.

How much does the Gemini 3.8 Flash API cost?

On the standard tier, Gemini 3.8 Flash costs $0.75 per 1M tokens for input and $3.75 per 1M tokens for output, as Google lists it.

What does Gemini 3.8 Flash take and return?

Gemini 3.8 Flash takes text, image, video, audio and returns text.

What is the knowledge cutoff of Gemini 3.8 Flash?

The knowledge cutoff of Gemini 3.8 Flash is March 2026.

Sources

Gemini 3.8 Flash | Gemini API | Google AI for Developers (read 16 October 2026)
Models | Gemini API | Google AI for Developers (read 16 October 2026)
Gemini deprecations | Gemini API | Google AI for Developers (read 16 October 2026)
Gemini 3.8 Flash - Model Card (read 16 October 2026)
Gemini Developer API pricing | Gemini API | Google AI for Developers (read 16 October 2026)

All of the Model Tracker