What is the Cheapest LLM API per Million Tokens Right Now? (Gemini Flash vs. DeepSeek V4 Flash vs. GPT-5.4 Nano)
Google Gemini 2.5 Flash-Lite and DeepSeek V4 Flash are currently in a dead heat for the crown of cheapest capable LLM API. Input tokens on these models start as low as $0.075 to $0.14 per million tokens, making them nearly 95% cheaper than flagship models like Claude 4.5 Sonnet or GPT-6.
Google Gemini 2.5 Flash-Lite: Priced at $0.075 per 1M input tokens and $0.30 per 1M output tokens. It features a massive 1M context window and near-instant time-to-first-token. It is ideal for document parsing and data extraction.
DeepSeek V4 Flash
Priced at $0.14 per 1M input tokens and $0.28 per 1M output tokens. It features dynamic off-peak discounts that slash prices by another 50% during non-working Asian business hours.
OpenAI GPT-5.4 Nano: Priced at $0.20 per 1M input and $0.80 per 1M output. While slightly more expensive than Flash-Lite, it features native JSON schema enforcement that rarely breaks formatting.
Cached Tokens Advantage
Both Google and DeepSeek offer aggressive prompt caching discounts (up to 75% off input tokens) if your system prompt or reference documents remain unchanged between calls.
RunAnAi Takeaway
For raw text processing and internal automation pipelines, route your traffic to Gemini Flash-Lite or DeepSeek V4 Flash. You can run hundreds of thousands of queries for the price of a sandwich.
Related Questions & User Queries
- • Cheapest AI API for developers in 2026
- • DeepSeek API pricing compared to Google Gemini Flash
- • Most affordable LLM API for high volume summarization
- • Lowest cost token pricing per million input and output