AI API Cost Over Time
Track how AI API prices have changed. Interactive charts showing historical pricing trends for OpenAI, Anthropic, Google, DeepSeek, and more.
Key Pricing Events
Major price changes and model launches that reshaped the AI API market.
DeepSeek V4 Pro launches at $0.44/$0.87
Near-Claude quality at 43% lower cost. Disrupts the mid-tier market and puts pressure on OpenAI and Anthropic pricing.
Claude 4 Opus/Sonnet retiring June 15
Anthropic deprecating Claude 4 Opus ($15/$75) and Sonnet 4 ($3/$15). Users migrating to Opus 4.8 ($5/$25) and Sonnet 4.6 ($3/$15).
GPT-5 mini drops to $0.25/$2.00
OpenAI's budget flagship undercuts Claude Haiku 4.5 by 75%. Best price-performance for general tasks.
xAI rebrands Grok 3 to Grok 4.3 at $1.25/$2.50
Major price cut from Grok 3's $3/$15. Now competitive with Gemini 2.5 Pro and GPT-5.
Google Gemini 2.0 Flash at $0.10/$0.40
Google's budget model becomes cheapest high-quality option. 93% cheaper than Claude Sonnet 4 with comparable output quality.
DeepSeek V4 Flash at $0.14/$0.28
Fastest and cheapest model with strong code performance. 1M context window. Becomes the default for high-volume production APIs.
GPT-5 launches at $1.25/$10.00
OpenAI's flagship drops 50% from GPT-4o's input price. Strong reasoning, 272K context. Sets new baseline for premium tier.
Claude Opus 4.8 at $5/$25
Anthropic's premium model drops 67% from Claude 4 Opus ($15/$75). Same quality, dramatically lower price.
Claude Sonnet 4 launches at $3/$15
Anthropic's mid-tier model. Strong coding and reasoning. Becomes the benchmark for quality-per-dollar.
GPT-4o drops to $2.50/$10.00
OpenAI cuts GPT-4o pricing by 50% as competition intensifies. Input drops from $5 to $2.50.
Gemini 1.5 Pro at $1.25/$5.00
Google's mid-tier model with 2M context. Competitive with GPT-4o at lower cost. Triggers industry-wide price competition.
The Pricing Trend Is Clear
AI API costs drop ~50% every 6 months. Budget models from Google and DeepSeek now cost less than premium models did 12 months ago. The "budget" tier of 2026 would have been "premium" in 2024.
What this means for you: If you're still on the same model you started with, you're likely overpaying by 40-80%. Use the Cost Optimizer to find your savings.
See how much you could save. Enter your current API setup and get a personalized cost optimization report.
Get Your Free Cost Report or Cost Calculator