Tokenova

Anthropic vs DeepSeek

Claude Opus 5 vs DeepSeek V4 Pro: pricing comparison

DeepSeek V4 Pro costs 20.1× less than Claude Opus 5 for a typical 1,000-input, 500-output request. Here is how the two compare on price, context and discounts.

Prices last updated:

Claude Opus 5DeepSeek V4 Pro
ProviderAnthropicDeepSeek
Input per 1M tokens$5.00$0.435
Output per 1M tokens$25.00$0.87
Context window1M1M
Batch discount50% offNot offered
Typical request (1,000 in / 500 out)$0.0175$0.00087
Monthly at 1,000 requests/day$525.00$26.10

DeepSeek V4 Pro: Open-weight model; price shown is DeepSeek's first-party API (cache-miss input).

Compare with your own numbers

Claude Opus 5 per request
$0.0175
Per month
$525.00
DeepSeek V4 Pro per request
$0.00087
Per month
$26.10

Cost by task

Cost per 1,000 requests at standard prices.
TaskTokens (in / out)Claude Opus 5DeepSeek V4 Pro
Short chat reply500 / 300$10.00$0.4785
RAG answer6,000 / 600$45.00$3.13
Summarize a long document12,000 / 1,000$85.00$6.09
Output-heavy generation1,000 / 4,000$105.00$3.92

Output tokens usually dominate the bill — see input vs output token cost.

Frequently asked questions

Is Claude Opus 5 or DeepSeek V4 Pro cheaper?
DeepSeek V4 Pro is cheaper for a typical request of 1,000 input and 500 output tokens: $0.00087 versus $0.0175 for Claude Opus 5, about 95% less.
What are the per-token prices of Claude Opus 5 and DeepSeek V4 Pro?
Claude Opus 5 costs $5.00 input and $25.00 output per million tokens. DeepSeek V4 Pro costs $0.435 input and $0.87 output per million tokens.
Which has the larger context window, Claude Opus 5 or DeepSeek V4 Pro?
Both accept 1M tokens of context.

More comparisons

All comparisons →