AI Model Pricing Comparison
Compare real-time pricing for major LLM APIs including OpenAI, Anthropic, Google, and Chinese AI providers like DeepSeek, Kimi, Doubao, and Qwen. Find the best value AI models with up-to-date token costs and comprehensive analysis.
🔄 Last updated: July 27, 2025 at 06:54 AM
💰
Real-time Pricing
Up-to-date token costs from official provider sources
📊
Easy Comparison
Side-by-side comparison of all major AI models
🔍
Smart Filtering
Sort and filter to find the perfect model for your needs
Current Pricing (per 1M tokens)
| Cached Input ($/M) | Max Context | Vision | ||||
|---|---|---|---|---|---|---|
Alibaba | Qwen 2.5 Turbo (1M) | — | $0.05 | $0.10 | 1M | Text Only |
OpenAI | GPT-4.1 nano | $0.025 | $0.10 | $0.40 | ~1M | Text Only |
Google | Gemini 2.5 Flash-Lite | $0.025 | $0.10 | $0.40 | 1M | 👁️ Vision |
Doubao | Doubao-1.5-pro (32k) | $0.022 | $0.11 | $0.28 | 32k | Text Only |
DeepSeek | DeepSeek-V3 Chat | $0.070 | $0.27 | $1.10 | 64k | Text Only |
Google | Gemini 2.5 Flash | $0.075 | $0.30 | $2.50 | 1M | 👁️ Vision |
Alibaba | Qwen 2.5 7B Instruct | — | $0.30 | $0.30 | 131k input/8k output | Text Only |
OpenAI | GPT-4.1 mini | $0.100 | $0.40 | $1.60 | ~1M | Text Only |
DeepSeek | DeepSeek-R1 Reasoner | $0.140 | $0.55 | $2.19 | 64k | Text Only |
OpenAI | GPT-4o mini | $0.300 | $0.60 | $2.40 | 128k | 👁️ Vision |
Moonshot AI | Kimi K2 | $0.150 | $0.60 | $2.50 | ~128k | Text Only |
Doubao | Doubao-1.5-pro (256k) | — | $0.69 | $1.24 | 256k | Text Only |
Alibaba | Qwen Chat 72B | — | $1.00 | $1.00 | 34k | Text Only |
Google | Gemini 2.5 Pro (≤200k) | $0.310 | $1.25 | $10.00 | 1M | 👁️ Vision |
OpenAI | GPT-4.1 | $0.500 | $2.00 | $8.00 | ~1M | Text Only |
Google | Gemini 2.5 Pro (>200k) | $0.625 | $2.50 | $15.00 | 1M | 👁️ Vision |
Anthropic | Claude 4 Sonnet | — | $3.00 | $15.00 | 200k | 👁️ Vision |
OpenAI | GPT-4o | $2.500 | $5.00 | $15.00 | 128k | 👁️ Vision |
Anthropic | Claude 4 Opus | — | $15.00 | $75.00 | 200k | 👁️ Vision |
*Pricing data updated July 2025. Cached prices shown where available (prompt caching reduces input costs).
Data compiled from official provider documentation and pricing pages. Some models use tiered pricing based on context length.