LLM API cost calculator
Estimate the monthly cost of DeepSeek, GLM, and Gemini APIs from your request volume, input and output tokens, cache hits, and DeepSeek peak hours.
Cost at official API prices
Per-million-token prices from the providers' pricing pages, checked on November 13, 2026. Open the price list before launch: providers change prices without notice.
| Model | Per month | Per 1,000 requests | Source |
|---|---|---|---|
| DeepSeek V4.1 Flash DeepSeek · $0.15 / $0.6 per 1M in / out Off-peak is half price; peak is 01:00–04:00 and 06:00–10:00 UTC on weekdays |
$6.00 | $0.60 | pricing |
| DeepSeek V4 Pro DeepSeek · $0.66 / $1.98 per 1M in / out |
$23.10 | $2.31 | pricing |
| GLM-5.3-Flash Z.ai · $0.15 / $0.5 per 1M in / out |
$5.50 | $0.55 | pricing |
| GLM-5.3 Z.ai · $1.4 / $4.4 per 1M in / out |
$50.00 | $5.00 | pricing |
| Gemini 3.8 Flash Google · $0.75 / $3.75 per 1M in / out Price through Dec 31, 2026; Google doubles it from Jan 1, 2027 ($1.50 / $7.50) |
$33.75 | $3.38 | pricing |
Prefer free limits? See the free API catalog and signup credits.
FAQ
How do I estimate the cost of an LLM API call?
Multiply input tokens (system prompt, history, question) and output tokens (answer) by their per-million prices and divide by one million. The calculator does this for your whole monthly volume.
What does prompt caching save?
When the start of a prompt repeats, providers bill it at the cache rate: about 50 times cheaper on DeepSeek and roughly 5–10 times cheaper on Gemini and GLM.
Are there fully free options?
Yes: OpenRouter free models, the Gemini API free tier, Groq, Cerebras, and others with request limits. See the free API catalog on the home page.