How Gemini 3.8 TTS pricing works
Gemini 3.8 TTS bills text input and generated audio output separately per one million tokens. Google documents a deterministic conversion of 25 audio tokens per second, so the calculator can convert generated minutes or hours into output tokens without inventing a speaking-speed assumption.
The duration totals are not flat bundled prices. They include the audio output rate for the selected mode plus the text input tokens you enter. Google does not offer context caching for these TTS models, so there is no cached-input discount or cache-storage charge to add.
For a worked explanation of the conversion and how the gap grows with volume, read the Gemini 3.8 Flash TTS pricing guide.
Flash vs Flash-Lite pricing
Google positions Gemini 3.8 Flash TTS for studio-grade fidelity, expressive acting, and long-form stability. It positions Flash-Lite for high-throughput, low-latency, cost-efficient conversational speech. The comparison block reports only the calculated price difference for the same workload; it does not turn those official descriptions into a quality ranking.
Example TTS workloads
Use 10 minutes for a short narration, one hour for a podcast-length output, 10 hours for a small audiobook or dubbing batch, and 100 hours for a scaled monthly production estimate. Monthly generations multiply both text input and generated audio costs.
Verify the data and continue planning
Review the Google Gemini API pricing page for all source-linked model records, inspect the AI API pricing dataset, or track the rollout in the Model Price Monitor. For ordinary text-output models, use the AI API Cost Calculator.
Frequently asked questions
How much does Gemini 3.8 Flash TTS cost per hour?
At the Standard audio-output rate through December 31, 2026, one hour uses 90,000 audio tokens and costs $0.81 before text-input cost. From January 1, 2027, the audio-output portion becomes $1.62 per hour.
How much does Gemini 3.8 Flash-Lite TTS cost per hour?
At the Standard audio-output rate through December 31, 2026, one hour costs $0.54 before text-input cost. From January 1, 2027, the audio-output portion becomes $1.08 per hour.
How are Gemini TTS audio tokens billed?
Google documents 25 audio output tokens per second. This calculator converts duration with that official rate, then adds separately billed text input tokens.
Is Flash-Lite cheaper than Flash TTS?
For the same mode and workload, Flash-Lite has a lower audio-output token rate. The text-input rates are the same. Price alone does not determine voice quality or product fit.
Does Gemini 3.8 TTS pricing change in 2027?
Yes. Google lists higher text input and audio output rates starting January 1, 2027. The calculator shows the current and future workload impact separately.