Gemini 3.8 Flash
The current public Gemini. Much faster than the frontier models, introductory pricing until the end of 2026, and a clear step below Argon on the index.
Intro price
41Index · High
$1.24Per measured task
33.1 pts/$Points / $
236Tokens / sec
Every published effort
| Setting | Index | Per task | Tokens / sec |
|---|---|---|---|
| High | 41 | $1.24 | 236 |
| Medium | 40 | $0.93 | — |
| Low | 33 | — | — |
What this workload costs
List price for one call. Reasoning tokens count as output. A cache hit uses the lab’s cache rate when one is published.
Gemini 3.8 Flash
Cache appliedSpecs
- API id
- gemini-3.8-flash
- Released
- 2 Sep 2026
- Context
- 1.05M
- Max output
- 66k
- Input / 1M
- $0.75
- Output / 1M
- $3.75
- Cache read / 1M
- $0.07
- Knowledge
- —
- Inputs
- Text, Image, Audio, Video
- Weights
- Closed
Introductory Standard price through 31 Dec 2026. From 1 Jan 2027 Google's pricing page lists $1.50 input and $7.50 output. Output includes thinking tokens. Batch is half.
Use it when
Multimodal product work you need to ship now: image, audio, and video in, with a 1M window and very fast output. High effort is 41 on the index at about 236 tokens per second.
Leave it when
As a stand-in for Argon, Opus, or Sol on the hardest agent tasks. The index gap is about ten points.