GoogleGemini 4 Argon
Google’s reasoning model for coding, agents, mathematics, and science. Input, reasoning, and output share a 1,048,576-token window. Synchronous output is capped at 65,536 tokens; background long decode can use the remaining window up to 1,048,576. Supports implicit caching only.
Source: Model specification, Long decode and thinking, Implicit cache threshold, Release notes, Approved OpenRouter offers
- Text
- Image
- Video
- Audio
- Document
- Text
Pricing
Every billing leg from the live catalog, in USD per 1M tokens.
| Billing leg | Rate | Batch |
|---|---|---|
| Input | $2 / M | $1 / M |
| Output | $10 / M | $5 / M |
| Cache read | $0.1 / M | $0.05 / M |
Cost calculator
Billing-grade math over the rate card above, including the thinking tokens most estimates miss.
Hidden reasoning tokens bill as output. 1.0x = thinking off. Medium-effort reasoning typically lands near 3x; heavy reasoning runs 6x to 8x. Set it to match your workload.
Compare providers
The same model, priced across every platform that serves it. Lowest combined in/out rate is flagged.
| Platform | Input / M | Output / M | Cache read / M | Batch | Access |
|---|---|---|---|---|---|
GoogleLowest | $2 | $10 | $0.1 | ✓ | First-party API |
Cursor | $2 | $10 | $0.1 | - | Direct |
Price history
Has this model gotten cheaper or more expensive?
More from Google
Other models from the same provider.
Stop estimating. Track it.
SuperPenguin meters every Gemini 4 Argon call your team makes, prices it with this exact rate card, and shows spend by feature, customer, and team.
Rates are generated from SuperPenguin's live pricing catalog.