# Gemini 3.8 Flash

Gemini 3.8 Flash is a model from Google, released on 2 Sep 2026. The context window is 1,048,576 tokens. The maximum output is 65,536 tokens. Pricing is $0.75 per million input tokens and $3.75 per million output tokens.

- Developer: Google
- Id: gemini-3.8-flash
- Reported: 2026-09-02T15:31:23.681Z
- Released: 2026-09-02T00:00:00.000Z
- Sighting: 2026-09-08T18:55:33.920Z
  - [evidence link](https://www.designarena.ai/leaderboard/gamedev)
- Status: Released
- Canonical: https://alexgetman.com/models/google/gemini-3-8-flash/

## Testing Gemini 3.8 Flash in an application
Sources checked: 2026-11-12

Google's [Gemini 3.8 Flash documentation](https://aistudio.google.com/docs/models/gemini-3.8-flash) describes the model for software engineering, agent workflows and enterprise tasks. Treat that positioning as a starting point for a trial. Measure the quality and response time of your own requests before choosing it for an interactive feature or a background job.

Use a small workload with clear acceptance checks: extract records into a schema, answer questions against supplied documents, or complete a tool-driven task. Keep source material and tools constant, and record malformed outputs, unsupported claims, failed calls and retries. The [API cost worksheet](/analysis/ai-api-pricing/#results) converts total spend into cost per accepted result.

**Match the model to the provider offer.** Google AI Studio and Vertex AI are access routes with their own account configuration and availability. Use the exact API identifier and the source links attached to your selected offer on this page. The [Gemini hub](/gemini/) collects related changes; compare their dates with the [official release notes](https://aistudio.google.com/docs/changelog) when investigating changed behavior.


## Questions

### What is the context window of Gemini 3.8 Flash?

The context window is 1,048,576 tokens. The maximum output is 65,536 tokens.

### How much does Gemini 3.8 Flash cost?

Pricing is $0.75 per million input tokens and $3.75 per million output tokens. Cached input is $0.075 per million.

### When was Gemini 3.8 Flash released?

On 2 Sep 2026.

- `gemini-3.8-flash` · batch · Context caching price: $0.50 / million token hours · / 1,000,000 tokens per hour (storage price)
- `gemini-3.8-flash` · flex · Context caching price: $0.50 / million token hours · / 1,000,000 tokens per hour (storage price)
- `gemini-3.8-flash` · priority · Context caching price: $0.50 / million token hours · / 1,000,000 tokens per hour (storage price)
- `gemini-3.8-flash` · standard · Context caching price: $0.50 / million token hours · / 1,000,000 tokens per hour (storage price)
## Timeline

- 2026-09-02T15:31:23.681Z · [Alex Getman](https://alexgetman.com/382/google-released-gemini-3-8-flash/)
> 🚨 Google released Gemini 3.8 Flash
> 
> • still $0.75 per 1M input tokens
> • at the same price, it is noticeably stronger than Gemini 3.7 Flash across almost the entire table
> • especially big gains in autonomous computer use and complex scientific tasks
> • beats GPT 5.6 Terra, a model in the same class, on almost every benchmark
- 2026-09-17T10:50:51.943Z · [Alex Getman](https://alexgetman.com/426/gemini-3-8-flash-just-appeared-on-arena-again/)
> 🚨 Gemini 3.8 Flash just appeared on Arena again
> 
> Looks like Google is testing a fresh checkpoint under that name.
> 
> My bet: either Gemini 3.9 Flash or an early Gemini 4.0 Pro checkpoint.
- 2026-09-24T17:49:05.746Z · [Alex Getman](https://alexgetman.com/465/right-now-you-can-test-5-models-for-free-in-cline/)
> Right now, you can test 5 models for free in Cline:
> 
> Gemini 3.8 Flash
> MiniMax M3.1 👀
> MiMo-V2.6-Flash
> DeepSeek V4.1 Flash
> Muse Spark 1.3
- 2026-10-07T19:11:03.647Z · [Gemini · deprecations](https://ai.google.dev/gemini-api/docs/deprecations?hl=en)
- 2026-10-07T19:11:03.647Z · [Gemini · deprecations](https://ai.google.dev/gemini-api/docs/deprecations?hl=en)
- 2026-10-09T09:05:16.967Z · [Gemini · API changelog](https://ai.google.dev/gemini-api/docs/changelog#10-08-2026)
- 2026-11-03T18:45:17.903Z · [Gemini · API changelog](https://ai.google.dev/gemini-api/docs/changelog#11-02-2026)
**Confirmed use**
- 2026-11-13T17:08:41.375Z · [Google · official articles · Introducing Gemini 3.8 Flash and 3.8 Flash Cyber](https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/)
> Gemini 3.8 Flash built this game with a simple prompt using a looping instruction in Google Antigravity
**Confirmed use**
- 2026-11-13T17:08:41.375Z · [Google · official articles · Introducing Gemini 3.8 Flash and 3.8 Flash Cyber](https://deepmind.google/blog/introducing-gemini-3-8-flash-and-38-flash-cyber/)
> Gemini 3.8 Flash built this game with a simple prompt using a looping instruction in Google Antigravity
**Confirmed use**
- 2026-11-13T19:49:14.434Z · [Google · official articles · See what 4 builders are making with Gemini 3.8 Flash](https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-flash-developers/)
> Ashutosh Shrivastava (@ai_for_success on X) used Gemini 3.8 Flash and Google Antigravity to map the live paths of satellites, space stations, and orbital rockets

## Evidence links

- [gemini-api-changelog](https://ai.google.dev/gemini-api/docs/changelog#09-02-2026)
- [Gemini · API changelog](https://ai.google.dev/gemini-api/docs/changelog#10-08-2026)
- [Gemini · API changelog](https://ai.google.dev/gemini-api/docs/changelog#11-02-2026)
- [Gemini · deprecations](https://ai.google.dev/gemini-api/docs/deprecations?hl=en)
- [review](https://aistudio.google.com/api-reference/models)
- [review](https://aistudio.google.com/docs/changelog)
- [gemini-model-index](https://aistudio.google.com/docs/models/gemini-3.8-flash)
- [gemini-pricing](https://aistudio.google.com/docs/pricing)
- [editorial](https://alexgetman.com/382/google-released-gemini-3-8-flash/)
- [editorial](https://alexgetman.com/426/gemini-3-8-flash-just-appeared-on-arena-again/)
- [editorial](https://alexgetman.com/465/right-now-you-can-test-5-models-for-free-in-cline/)
- [Arena](https://arena.ai)
- [Artificial Analysis · benchmarks](https://artificialanalysis.ai/models)
- [Arena · leaderboards](https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/)
- [Google · official articles · See what 4 builders are making with Gemini 3.8 Flash](https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-flash-developers/)
- [vertex-google-models](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemini-3.8-flash)
- [Google · official articles · Introducing Gemini 3.8 Flash and 3.8 Flash Cyber](https://deepmind.google/blog/introducing-gemini-3-8-flash-and-38-flash-cyber/)
- [models.dev · catalogue](https://models.dev/#gemini-3-8-flash)
- [models.dev · catalogue](https://models.dev/#gemini-3.8-flash)
- [OpenCode · discovery](https://opencode.ai/data/google/gemini-3-8-flash)
- [OpenRouter](https://openrouter.ai/google/gemini-3.8-flash)
- [VoxelBench · leaderboard](https://voxelbench.ai/leaderboard)
- [designarena:gamedev](https://www.designarena.ai/leaderboard/gamedev)
- [DesignArena · uicomponent](https://www.designarena.ai/leaderboard/ui-components)
- [DesignArena · website](https://www.designarena.ai/leaderboard/website)
- [DesignArena · gamedev](https://www.designarena.ai/leaderboard?category=gamedev)
- [DesignArena · uicomponent](https://www.designarena.ai/leaderboard?category=uicomponent)
- [DesignArena · website](https://www.designarena.ai/leaderboard?category=website)

## Signals

- [Gemini 3.8 Flash spotted on DesignArena](https://alexgetman.com/signals/2026/09/08/gemini-3-8-flash-high/)