Skip to content

Vertex AI Generative AI Pricing

Complete pricing guide for building and deploying AI models in Vertex AI. Prices are in USD.

Pricing Overview

Billing Note

You're charged only for requests that return a 200 response code. Requests returning 4xx and 5xx codes aren't charged.


Google Models

Gemini 3 Pro Preview

TypePrice /1M tokens (≤200K)Price /1M tokens (>200K)Cached (≤200K)Batch API
Input (text, image, video, audio)$2$4$0.20$1
Text Output (response + reasoning)$12$18N/A$6
Image Output$120N/AN/A$60

Image Output Tokens:

  • 1K/2K image (1024x1024, 2048x2048): 1,120 tokens = ~$0.134/image
  • 4K image (4096x4096): 2,000 tokens = ~$0.24/image

Gemini 2.5 Pro

TypePrice /1M tokens (≤200K)Price /1M tokens (>200K)CachedBatch API
Input (text, image, video, audio)$1.25$2.50$0.125$0.625
Text Output (response + reasoning)$10$15N/A$5

Gemini 2.5 Flash

TypePrice /1M tokensCachedBatch API
Input (text, image, video)$0.30$0.030$0.15
Audio Input$1.00$0.10$0.50
Text Output$2.50N/A$1.25
Image Output$30N/A$15

Gemini 2.5 Flash Lite

TypePrice /1M tokensCachedBatch API
Input (text, image, video)$0.10$0.010$0.05
Audio Input$0.30$0.030$0.15
Text Output$0.40N/A$0.20

Gemini 2.0 Flash

TypePrice /1M tokensBatch API
Input tokens$0.15$0.075
Input audio tokens$1.00$0.50
Output text tokens$0.60$0.30
Tuning (training tokens)$3.00-

Gemini 2.0 Flash Lite

TypePrice /1M tokensBatch API
Input tokens$0.075$0.0375
Input audio tokens$0.075$0.0375
Output text tokens$0.30$0.15
Tuning (training tokens)$1.00-

Gemini Model Comparison


Grounding Services

ServiceFree TierPrice
Google Search (Flash models)1,500 prompts/day combined$35/1,000 prompts
Google Search (Pro models)10,000 prompts/day$35/1,000 prompts
Web Grounding EnterpriseNone$45/1,000 prompts
Grounding with Your DataNone$2.50/1,000 requests
Google MapsVaries by model$25/1,000 prompts

Gemini 3 Pro Grounding

Gemini 3 Pro includes 5,000 search queries/month free. Excess billed at $14/1,000 queries. Billing starts January 5, 2026.


Imagen Models

ModelFeaturePrice
Imagen 4 UltraImage generation$0.06/image
Imagen 4Image generation$0.04/image
Imagen 4Upscaling (2K-4K)$0.06/image
Imagen 4 FastImage generation$0.02/image
Imagen 3Image generation/editing$0.04/image
Imagen 3 FastImage generation$0.02/image
Imagen 2/1Generation/editing$0.02/image
ImagenVisual Captioning$0.0015/image
ImagenVisual Q&A$0.0015/image
ImagenProduct Recontext$0.12/image
Virtual Try-OnClothing visualization$0.06/image

Veo Video Models

ModelFeatureResolutionPrice
Veo 3.1Video + Audio720p/1080p$0.40/sec
Veo 3.1Video only720p/1080p$0.20/sec
Veo 3.1 FastVideo + Audio720p/1080p$0.15/sec
Veo 3.1 FastVideo only720p/1080p$0.10/sec
Veo 3Video + Audio720p/1080p$0.40/sec
Veo 3Video only720p/1080p$0.20/sec
Veo 3 FastVideo + Audio720p/1080p$0.15/sec
Veo 3 FastVideo only720p/1080p$0.10/sec
Veo 2Video generation720p$0.50/sec

Lyria Music Generation

ModelFeaturePrice
Lyria 2Music generation$0.06/30 seconds

Embedding Models

ModelTypeOnline Price /1K tokensBatch Price /1K tokens
Gemini EmbeddingInput$0.00015$0.00012
Text Embeddings (non-Gemini)Input$0.000025$0.00002
multilingual-e5-smallInput$0.000015$0.0000075
multilingual-e5-largeInput$0.000025$0.0000125

Multimodal Embeddings

Input TypePrice
Text$0.0002/1k chars
Image$0.0001/image
Video Plus$0.002/sec (15 embeddings/min)
Video Standard$0.001/sec (8 embeddings/min)
Video Essential$0.0005/sec (4 embeddings/min)

Partner Models

Anthropic Claude Models

ModelInput /1MOutput /1MBatch InputBatch Output
Claude Opus 4.5$5.00$25.00$2.50$12.50
Claude Sonnet 4.5$3.00$15.00$1.50$7.50
Claude Haiku 4.5$1.00$5.00$0.50$2.50
Claude Opus 4.1$15.00$75.00$7.50$37.50
Claude Sonnet 4$3.00$15.00$1.50$7.50
Claude 3.5 Haiku$0.80$4.00$0.40$2.00
Claude 3 Haiku$0.25$1.25--

Claude Web Search: $10/1,000 searches

Meta Llama Models

ModelInput /1MOutput /1MBatch InputBatch Output
Llama 3.1 405B$5.00$16.00--
Llama 3.3 70B$0.72$0.72$0.36$0.36
Llama 4 Scout$0.25$0.70$0.125$0.35
Llama 4 Maverick$0.35$1.15$0.175$0.575

DeepSeek Models

ModelInput /1MOutput /1MBatch InputBatch Output
DeepSeek-V3.1$0.60$1.70$0.30$0.85
DeepSeek-R1 (0528)$1.35$5.40$0.675$2.70

Qwen Models

ModelInput /1MOutput /1MBatch InputBatch Output
Qwen3-Coder-480B$1.00$4.00$0.50$2.00
Qwen3-235B$0.25$1.00$0.125$0.50
Qwen3-Next-80B-Thinking$0.15$1.20--
Qwen3-Next-80B-Instruct$0.15$1.20--

Mistral AI Models

ModelInput /1MOutput /1M
Mistral Medium 3$0.40$2.00
Mistral Small 3.1$0.10$0.30
Codestral 2$0.30$0.90

OpenAI Models on Vertex

ModelInput /1MOutput /1MBatch InputBatch Output
gpt-oss-120b$0.09$0.36$0.045$0.18
gpt-oss-20b$0.07$0.25$0.035$0.125

Provisioned Throughput

For guaranteed throughput using Generative AI Scale Units (GSUs):

CommitmentPrice per GSU
1 week$1,200/week
1 month$2,700/month
3 months$2,400/month
1 year$2,000/month

Example Cost Calculation

Scenario: 10 QPS with 1,000 text tokens + 500 audio tokens input, 300 text tokens output using Gemini 2.0 Flash

Burndown calculation:
- Input text: 1,000 × 1 = 1,000 tokens
- Input audio: 500 × 7 = 3,500 tokens
- Output text: 300 × 4 = 1,200 tokens
- Total per query: 5,700 tokens
- At 10 QPS: 57,000 tokens/sec
- GSUs needed: 57,000 ÷ 3,360 = ~17 GSUs

Cost:
- 1 week: $1,200 × 17 = $20,400
- 1 month: $2,700 × 17 = $45,900
- 1 year: $2,000 × 17 = $34,000/month

Model Tuning Pricing

ModelSupervised Fine-tuning /1M tokens
Gemini 2.5 Pro$25
Gemini 2.5 Flash$5
Gemini 2.5 Flash Lite$1.50
Gemma 3 27B IT$6.83
Llama 3.1 8B$0.67
Llama 3.3 70B$6.72
Llama 4 Scout$5.77
Qwen 3 32B$6.57

Training Token Calculation

Training tokens = (tokens in dataset) × (number of epochs)


Context Cache Storage

ModelStorage Price /M tokens/hour
Gemini 3 Pro$4.50
Gemini 2.5 Pro$4.50
Gemini 2.5 Flash$1.00
Gemini 2.5 Flash Lite$1.00
Gemini 2.0 Flash$1.00
Gemini 2.0 Flash Lite$1.00

Quick Reference: Best Value Models

Use CaseRecommended ModelInput Cost
High-volume chatbotsGemini 2.0 Flash Lite$0.075/M
General purposeGemini 2.0 Flash$0.15/M
Advanced reasoningGemini 2.5 Pro$1.25/M
Image generation (fast)Imagen 4 Fast$0.02/image
Video generationVeo 3 Fast$0.10/sec
EmbeddingsText Embeddings$0.000025/1k chars

Last updated: December 2025

Built with VitePress & Mermaid diagrams