Gemini 3.5 Flash vs Gemini 3.8 Flash

Wins 0 of 5 benchmarks
$0.93 / task
Wins 5 of 5 benchmarks
Humanity's Last Exam Humanity's Last Exam: Very hard expert questions across many fields, written to stump today’s models
Gemini 3.5 Flash41.3
Gemini 3.8 Flash42.1
Terminal-Bench 4.0 Terminal-Bench 4.0 · source setting 2: Agents use a terminal to complete executable coding, system administration and research tasks in the 4.0 suite
Gemini 3.5 Flash · high6.6
Gemini 3.8 Flash19.7
AutomationBench-AA AutomationBench-AA: Agents orchestrate SaaS workflows through APIs; objectives receive credit only when the task has no guardrail violation
Gemini 3.5 Flash · high42.1
Gemini 3.8 Flash60.9
Arena Text · creative writing Arena Text · creative writing: Human pairwise preferences in the stated text vote pool
Gemini 3.5 Flash1,468
Gemini 3.8 Flash · high1,487
MMMU-Pro (vision) MMMU-Pro (vision): College-level questions that require understanding images: charts, diagrams, photos
Gemini 3.5 Flash83.9
Gemini 3.8 Flash84.2

Specs

SpecGemini 3.5 FlashGemini 3.8 Flash
Input, per 1M$1.50!$0.75
Output, per 1M$9!$3.75
Cost per task?—$0.93
Speed, tok/s199—
Context1M1M
WeightsClosedClosed
ReleasedMay 2026Sep 2026
See images✓✓
Read PDFsPartial✓
Hear audio✓✓
Watch video✓✓
Call tools✓✓
Force a tool callPartialPartial
Return strict JSON✓✓
Search the web?✓
Run code?✓
Use a computer??

What breaks when you switch

Gemini 3.5 Flash → Gemini 3.8 Flash
Sampling parameters are ignored›

temperature, top_p and top_k are ignored now and will cause HTTP 400 in future generations; remove them from requests.

No minimal thinking level›

The minimal thinking level returns an error; avoid minimal.

Gemini 3.8 Flash → Gemini 3.5 Flash
Nothing known.