DeepSeek Token Counter
Estimate DeepSeek token counts for your prompts — and price them with DeepSeek's official API rates, not third-party reseller prices. Off-peak discounts and 98%-off cache reads included.
This is a character-based heuristic (~3.5 characters per token for English), not DeepSeek's actual tokenizer. Treat it as a rough estimate, not a bill.
What this text would cost, per model
Sorted cheapest first. Assumes 0 input tokens and 500 output tokens. Counts for non-OpenAI rows are estimates.
| Model | Input / 1M | Output / 1M | Est. cost | With cache hits |
|---|---|---|---|---|
| Gemini 2.5 Flash-Lite · Google | $0.1 | $0.4 | $0.0002 | — |
| GPT-6 Luna · OpenAI | $0.1 | $0.5 | $0.00025 | $0.00025 |
| DeepSeek-V4.1 Flash · DeepSeek | $0.15 | $0.6 | $0.0003 | $0.0003 |
| DeepSeek-V4 Pro · DeepSeek | $0.66 | $1.98 | $0.00099 | — |
| Gemini 3.8 Flash · Google | $0.75 | $3.75 | $0.001875 | — |
| Claude Haiku 4.5 · Anthropic | $1 | $5 | $0.0025 | $0.0025 |
| GPT-6 Sol · OpenAI | $2 | $10 | $0.005 | $0.005 |
| Claude Sonnet 5 · Anthropic | $2 | $10 | $0.005 | $0.005 |
| Claude Sonnet 5.5 · Anthropic | $2 | $10 | $0.005 | — |
| Gemini 3.1 Pro Preview · Google | $2 | $12 | $0.006 | — |
| Claude Opus 5.5 · Anthropic | $4 | $20 | $0.01 | $0.01 |
| Claude Opus 5 · Anthropic | $5 | $25 | $0.0125 | $0.0125 |
| GPT-6 Astra · OpenAI | $10 | $50 | $0.025 | — |
| Claude Fable 5.1 · Anthropic | $10 | $50 | $0.025 | $0.025 |
“With cache hits” assumes all input tokens hit the prompt cache, where the provider publishes a cached-input rate. DeepSeek rows use off-peak rates; peak hours cost 2×.
Prices change frequently — last verified 2026-09-30. Rates below are USD per 1M tokens from public pricing pages. Always confirm on the provider's official pricing page before budgeting.
Official pricing, off-peak shown. DeepSeek's own API rates are shown here — some tools quote higher third-party reseller rates instead. Off-peak traffic is discounted; peak hours cost 2×. Cached input is billed at 2% of the input rate.
About DeepSeek token counting
DeepSeek does not ship a public client-side tokenizer, so the count above is a transparent heuristic estimate. What we can be precise about is price: the table uses DeepSeek's published rates (DeepSeek-V4.1 Flash and DeepSeek-V4 Pro), with the off-peak discount and cache-read rate spelled out instead of hidden.
Comparing providers? The main LLM token counter puts DeepSeek side by side with OpenAI, Claude, and Gemini, and the full API pricing table lists every model's rates.
Frequently asked questions
Is the DeepSeek token count exact?
No — DeepSeek has not published a standalone client-side tokenizer, so this page shows a clearly-labeled character-based estimate (~3.5 characters per token for English). Code and non-English text can deviate, so treat it as a rough guide for budgeting, not a bill.
Why do you use DeepSeek's official pricing instead of reseller rates?
Some tools quote third-party reseller prices (for example, API aggregator rates), which are higher than DeepSeek's own API. This page uses DeepSeek's official published rates, so the comparison reflects what you would actually pay calling DeepSeek directly.
What are off-peak and peak rates?
DeepSeek discounts API traffic during off-peak hours: the rates shown here are off-peak, and peak-hour traffic costs 2×. If your workload runs on a schedule, shifting batch jobs to off-peak hours roughly halves the bill.
How much does prompt caching save on DeepSeek?
Cached input is billed at just 2% of the input rate — a 98% discount, the steepest among major providers. For agents that resend long context every turn, cache hits dominate the total cost.
Does my text leave my browser?
No. Counting and cost math run locally in JavaScript — nothing is uploaded or logged.