Both models have prices that move: DeepSeek V4-Pro by time of day, Gemini 4 Argon by date. Which is cheaper depends on when you send traffic — but V4-Pro stays the lower-cost option in every combination.
Off-peak, DeepSeek V4-Pro is about 3× cheaper on input and 5× on output than Gemini 4 Argon at its intro rate. During DeepSeek's peak hours, or after Argon doubles on December 1, the ratios shift — but V4-Pro remains cheaper. Argon is a flagship with image input; V4-Pro is a text-only mid-tier model.
| DeepSeek V4-Pro | Gemini 4 Argon | |
|---|---|---|
| Provider | DeepSeek | |
| Tier | Mid-tier | Flagship |
| Input / 1M tokens | $0.66 | $2.00 |
| Output / 1M tokens | $1.98 | $10.00 |
| Cache hit / 1M | $0.022 | $0.10 |
| Cache write / 1M | — | — |
| Context window | 1M | 1M |
| Notes | A planned Sep 14, 2026 reroute to V4.1-Flash rates was reversed after user pushback; billing unchanged since the Aug 13, 2026 GA build. Off-peak rate shown; peak (01:00-04:00 & 06:00-10:00 UTC weekdays) is 2x: $1.32/$3.96 | New top frontier model, generally available Oct 13, 2026 for AI Pro/Ultra subscribers and via the API (model id gemini-4-argon); long-decode generation up to 1M tokens; intro price thru Nov 30, 2026, then $4/$20 (cache read $0.20); not available on the free tier |
DeepSeek V4-Pro is 3.0× cheaper on input tokens ($0.66 vs $2.00 per 1M). DeepSeek V4-Pro is 5.1× cheaper on output tokens ($1.98 vs $10.00 per 1M).
| Workload | DeepSeek V4-Pro | Gemini 4 Argon | Difference |
|---|---|---|---|
| Support chatbot 10,000 conversations × (2,000 input + 400 output tokens) | $21.12 | $80.00 | DeepSeek V4-Pro saves $58.88 |
| Document Q&A 1,000 questions × (30,000 input + 800 output tokens) | $21.38 | $68.00 | DeepSeek V4-Pro saves $46.62 |
| Coding agent session 4M input tokens (90% cache hits) + 150,000 output tokens | $0.64 | $2.66 | DeepSeek V4-Pro saves $2.02 |
Standard rates per 1M tokens from data on this site. The coding-agent row assumes 90% of input is read from the prompt cache (full input price where a model has no cache rate). Cache-write fees, batch discounts and long-context surcharges listed in the notes are not included. Updated November 2, 2026.
your workload is text-only, high-volume, and can be scheduled outside DeepSeek's weekday peak window.
you need a flagship-tier Google model, image or multimodal inputs, or prices that don't change with the hour.
DeepSeek doubles V4-Pro's rates from 01:00–04:00 and 06:00–10:00 UTC on weekdays. Google's Argon rate of $2.00/$10.00 lasts through November 30, 2026, then doubles. Peak-hour V4-Pro against intro-rate Argon is the closest the two get on output; off-peak V4-Pro against post-December Argon is the widest gap.
Argon applies cache pricing implicitly to repeated prefixes ($0.10 now) with no write fee listed. V4-Pro's cache hit is $0.022 off-peak. Both are low enough that cache-heavy agents are cheap on either; the uncached and output tokens decide the bill.
DeepSeek lists V4-Pro as text-only, with up to 384K output tokens per response. Both offer a 1M-token context window. Argon is not on the Gemini API free tier; Google's Batch API halves its rate for async jobs.
Yes, at both DeepSeek's off-peak and peak rates, and before and after Argon's December 1 price increase.
Google lists the intro rate through November 30, 2026, with prices doubling from December 1.
DeepSeek lists vision as not supported on V4-Pro. Gemini 4 Argon is Google's multimodal flagship.
How we choose: prices come from each vendor's official pricing page and are re-checked every week. We take no affiliate fees, so rankings aren't for sale.