Gemini API pricing
As of October 21, 2026, Google's Gemini and Gemma API prices run from $0.075 to $12.00 per million output tokens across 30 models, as listed on OpenRouter. Gemma 3 4B is the cheapest paid model at $0.045 input and $0.075 output per million tokens, and Nano Banana Pro (Gemini 3 Pro Image) the most expensive at $2.00 input and $12.00 output.
Gemini 4 Argon ranks first among Google models in our Vibe Coding Index, at 81.0 on 5 benchmarks. 11 of 30 Google models have any benchmark score. Cost per solved task is known for 10 of 30 models. 2 models are free.
| Model | Input $/Mtok | Output $/Mtok | Context | Evidence | Index | Cost per solved task |
|---|---|---|---|---|---|---|
| Gemini 4 Argon | $2.00 | $10.00 | 1.05M | 81.0 | No run cost | |
| Nano Banana 2.1 | $1.50 | $7.50 | 66K | Not benchmarked | No run cost | |
| Gemini 3.8 Flash | $0.75 | $3.75 | 1.05M | 61.6 | $2.74 | |
| Gemini 3.7 Flash | $0.75 | $3.75 | 1.05M | 58.0 | $1.78 | |
| Gemini 3.5 Flash Lite | $0.30 | $2.50 | 1.05M | 49.0 | $0.41 | |
| Gemini 3.6 Flash | $0.75 | $3.75 | 1.05M | 62.5 | $1.49 | |
| Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image) | $0.25 | $1.50 | 66K | Not benchmarked | No run cost | |
| Nano Banana 2 (Gemini 3.1 Flash Image) | $0.50 | $3.00 | 131K | Not benchmarked | No run cost | |
| Nano Banana Pro (Gemini 3 Pro Image) | $2.00 | $12.00 | 131K | Not benchmarked | No run cost | |
| Gemini 3.5 Flash | $1.50 | $9.00 | 1.05M | 39.8 | $1.21 | |
| Gemini 3.1 Flash Lite | $0.25 | $1.50 | 1.05M | Not benchmarked | No run cost | |
| Gemma 4 26B A4B | $0.15 | $0.60 | 262K | Not enough data | No run cost | |
| Gemma 4 31B | $0.14 | $0.40 | 262K | 39.0 | No run cost | |
| Lyria 3 Clip Preview | Free | Free | 1.05M | Not benchmarked | No run cost | |
| Lyria 3 Pro Preview | Free | Free | 1.05M | Not benchmarked | No run cost | |
| Gemini 3.1 Flash Lite Preview | $0.25 | $1.50 | 1.05M | 22.7 | $0.18 | |
| Nano Banana 2 (Gemini 3.1 Flash Image Preview) | $0.50 | $3.00 | 66K | Not benchmarked | No run cost | |
| Gemini 3.1 Pro Preview Custom Tools | $2.00 | $12.00 | 1.05M | Not benchmarked | No run cost | |
| Gemini 3.1 Pro Preview | $2.00 | $12.00 | 1.05M | 32.3 | $0.99 | |
| Gemini 3 Flash Preview | $0.50 | $3.00 | 1.05M | 43.7 | $0.37 | |
| Nano Banana Pro (Gemini 3 Pro Image Preview) | $2.00 | $12.00 | 66K | Not benchmarked | No run cost | |
| Nano Banana (Gemini 2.5 Flash Image) | $0.30 | $2.50 | 33K | Not benchmarked | No run cost | |
| Gemini 2.5 Flash Lite | $0.10 | $0.40 | 1.05M | Not benchmarked | No run cost | |
| Gemini 2.5 Flash | $0.30 | $2.50 | 1.05M | Not benchmarked | $0.466 | |
| Gemini 2.5 Pro | $1.25 | $10.00 | 1.05M | 19.1 | $0.64 | |
| Gemini 2.5 Pro Preview 06-05 | $1.25 | $10.00 | 1.05M | Not benchmarked | No run cost | |
| Gemma 3 12B | $0.075 | $0.1625 | 131K | Not benchmarked | No run cost | |
| Gemma 3 4B | $0.045 | $0.075 | 131K | Not benchmarked | No run cost | |
| Gemma 3 27B | $0.08 | $0.30 | 131K | Not benchmarked | No run cost | |
| Gemma 2 27B | $0.65 | $0.65 | 8K | Not benchmarked | No run cost |
| Model | Released | Retired | Last input $/Mtok | Last output $/Mtok |
|---|---|---|---|---|
| Gemini 2.0 Flash-Lite | February 25, 2025 | June 1, 2026 | $0.075 | $0.30 |
| Gemini 2.0 Flash | February 5, 2025 | June 1, 2026 | $0.10 | $0.40 |
| Gemini 1.5 Flash | May 23, 2024 | September 29, 2025 | $0.075 | $0.30 |
| Gemini 1.5 Pro | May 23, 2024 | September 29, 2025 | $1.25 | $5.00 |
Price changes
No change in a maker's own API price recorded yet. We read prices daily from OpenRouter (since October 15, 2026) and list a change only when the maker's own price moves between two of our snapshots.
Google API pricing questions
- How much does the Gemini API cost?
- Google's Gemini and Gemma API charges per token, with input and output priced separately. On October 21, 2026, Gemma 3 4B cost $0.045 input and $0.075 output per million tokens and Nano Banana Pro (Gemini 3 Pro Image) cost $2.00 input and $12.00 output, as listed on OpenRouter.
- Which Gemini model is best for coding?
- Gemini 4 Argon ranks first among Google models in our Vibe Coding Index, at 81.0 on 5 benchmarks. The index combines public coding benchmarks. We do not run our own.
- What does a working Gemini answer cost?
- Gemini 3.1 Flash Lite Preview costs $0.18 per solved SWE-bench task, the lowest known for Google. Cost per solved task is known for 10 of 30 models.
- Has Google changed its API prices?
- We have no recorded change. Price tracking started on October 15, 2026, and a change appears on this page the day a snapshot shows one.
- Is this the price of the Gemini subscription?
- No. These are API prices, billed per token. Consumer and team subscriptions are billed per seat and are not in our data.