4 models across 2 tiers

Gemini API pricing in 2026

Google has models in the mid and fast tiers: mid-tier (Gemini 3.1 Pro) and fast (Gemini 3.5 Flash, 3 Flash, 3.5 Flash-Lite). All four Gemini models list under $2.00/MTok input. GPT 5.6 Luna took the outright budget floor in August 2026. Google has a 90% cache read discount with no cache write surcharge, which makes it the cheapest family to maintain a prefix that changes often.

Models

Cheapest input

$0.30/MTok

Cache discount

90% off

Max context

1,000K

All Gemini models

Model	Tier	Input $/MTok	Output $/MTok	Context	Best task	Released
Gemini 3.5 Flash-Lite	Fast / Budget	$0.30	$2.50	1,000K	Classification (85/100)	Jun 2026
Gemini 3 Flash	Fast / Budget	$0.50	$3.00	1,000K	Classification (88/100)	Mar 2026
Gemini 3.5 Flash	Fast / Budget	$1.50	$9.00	1,000K	Classification (92/100)	May 2026
Gemini 3.1 Pro	Mid-tier	$2.00	$12.00	1,000K	Reasoning (90/100)	Apr 2026

Gemini cost by use case

Estimated cost per API call across common workloads, using default token counts for each use case.

Use case	Flash-Lite	Flash	Flash	Pro
Code Generation	$0.0106	$0.0130	$0.0390	$0.0520
Unit Test Generation	$0.0134	$0.0165	$0.0495	$0.0660
SQL Generation	$0.002950	$0.003750	$0.0113	$0.0150
API Generation	$0.0159	$0.0195	$0.0585	$0.0780
Code Refactoring	$0.0112	$0.0140	$0.0420	$0.0560
Code Review	$0.006500	$0.008500	$0.0255	$0.0340
PR Review	$0.009900	$0.0130	$0.0390	$0.0520
Security Review	$0.008050	$0.0105	$0.0315	$0.0420
Bug Detection	$0.006500	$0.008500	$0.0255	$0.0340
Document Summarization	$0.005500	$0.008000	$0.0240	$0.0320
PDF Summarization	$0.008250	$0.0120	$0.0360	$0.0480
Meeting Notes	$0.007400	$0.0100	$0.0300	$0.0400
Email Summarization	$0.002150	$0.003000	$0.009000	$0.0120
Chatbot	$0.003100	$0.004000	$0.0120	$0.0160
RAG Pipeline	$0.005550	$0.007500	$0.0225	$0.0300
Customer Support Bot	$0.003400	$0.004500	$0.0135	$0.0180
FAQ Bot	$0.002450	$0.003150	$0.009450	$0.0126
Data Extraction	$0.003700	$0.005000	$0.0150	$0.0200
JSON Extraction	$0.002900	$0.003900	$0.0117	$0.0156
Invoice Parsing	$0.002100	$0.002800	$0.008400	$0.0112
Resume Parsing	$0.002750	$0.003650	$0.0110	$0.0146
Chain of Thought	$0.0134	$0.0165	$0.0495	$0.0660
Legal Analysis	$0.0130	$0.0170	$0.0510	$0.0680
Math Reasoning	$0.007950	$0.009750	$0.0292	$0.0390
Financial Analysis	$0.0124	$0.0160	$0.0480	$0.0640
Content Moderation	$0.000400	$0.000550	$0.001650	$0.002200
Sentiment Analysis	$0.000400	$0.000550	$0.001650	$0.002200
Categorization	$0.000615	$0.000850	$0.002550	$0.003400
Intent Detection	$0.000340	$0.000450	$0.001350	$0.001800
Copywriting	$0.005150	$0.006250	$0.0188	$0.0250
Blog Writing	$0.0203	$0.0245	$0.0735	$0.0980
Marketing Copy	$0.007740	$0.009400	$0.0282	$0.0376
Product Descriptions	$0.003900	$0.004750	$0.0143	$0.0190

Gemini quality benchmarks by task

Confidence scores (0-100) across all task types. Higher is better.

Confidence scores are a third-party benchmark snapshot from week 29 of 2026. This snapshot is 8 weeks old, so treat the scores as directional and validate against your own evals.

Task	Flash-Lite	Flash	Flash	Pro
Code Generation	55/100	65/100	68/100	86/100
Code Review	52/100	62/100	65/100	84/100
Summarization	79/100	83/100	85/100	87/100
Q&A	78/100	82/100	84/100	87/100
Extraction	83/100	86/100	90/100	85/100
Reasoning	50/100	58/100	60/100	90/100
Classification	85/100	88/100	92/100	86/100
Creative Writing	45/100	52/100	55/100	84/100

Recent Gemini pricing changes

Gemini 3.5 Flash input increased

$0.15 to $1.50/MTok

Jul 15, 2026

Gemini 3.5 Flash output increased

$0.60 to $9.00/MTok

Jul 15, 2026

Gemini 3.5 Flash output decreased

$0.75 to $0.60/MTok

Jun 10, 2026

Gemini 3.1 Pro input launched

$2.00/MTok at launch

Apr 22, 2026

View full Gemini pricing history

Head-to-head comparisons

Track your Gemini spend

See exactly how much you spend on Gemini models and where you can save. Two-line wrapper install, no API keys shared.

Free spend teardown Start free dashboard