Updated 27 March 2026

Gemini API Pricing

Gemini Flash is the cheapest mainstream AI model: $0.075/M input tokens. Gemini Pro competes with GPT-4o and Claude Sonnet at slightly lower prices. All models have generous free tiers.

API Pricing by Model

ModelInputOutputContextBest ForFree Tier
Gemini 2.0 Flash$0.10/M$0.40/M1M tokensHigh-volume, low-cost tasks. Chat, classification, extraction.Yes (generous free tier)
Gemini 1.5 Flash$0.075/M$0.30/M1M tokensBudget production workloads. Good quality at lowest cost.Yes
Gemini 1.5 Flash-8B$0.0375/M$0.15/M1M tokensHighest throughput, lowest cost. Simple tasks only.Yes
Gemini 1.5 Pro$3.50/M$10.50/M2M tokensComplex reasoning, long documents, multimodal analysis.Limited
Gemini 2.5 Pro$1.25/M$10.00/M1M tokensLatest reasoning model. Best quality for complex tasks.Limited

Gemini vs OpenAI vs Claude API Costs

TaskGeminiOpenAIClaudeCheapest
Simple chat (1K tokens)$0.0005 (Flash)$0.003 (GPT-4o-mini)$0.003 (Haiku)Gemini Flash
Complex reasoning (4K tokens)$0.056 (2.5 Pro)$0.060 (GPT-4o)$0.075 (Sonnet)Gemini 2.5 Pro
Long document (100K context)$0.35 (1.5 Pro)$1.25 (GPT-4o)$0.75 (Sonnet)Gemini 1.5 Pro
Batch processing (1M requests)$75 (Flash)$1,500 (4o-mini)$750 (Haiku)Gemini Flash

Gemini's advantage: cost and context window.

Gemini Flash is 10-20x cheaper than GPT-4o-mini for simple tasks. Gemini 1.5 Pro offers a 2M token context window, the largest of any major model, at competitive prices. For long-document analysis, RAG over large codebases, or high-volume batch processing, Gemini is hard to beat on price. The trade-off: Claude and GPT-4o are generally regarded as higher quality for nuanced reasoning and creative writing.