Tools & price references
Gemini API pricing
Check Gemini Developer API prices across model types, then estimate text usage with prompt tiers and announced rate changes.
Free to useNo sign-upSources and formulas
Updated
Jump to the full model price reference
Estimated monthly cost
$80.00
USD / month10,000 requests · Gemini 4 Argon
- Uncached input
- $30.00
- Cached input
- $0.00
- Output
- $50.00
- Other monthly costs
- $0.00
- API / request
- $0.0080
Standard input and output rates, no caching.
Estimate, not a quote. Excludes taxes, non-text usage, tool fees and charges you have not entered.
Official model price reference
36 named model entries across 36 pricing rows. Source checked . Calculator presets above remain a selected set.
All 36 named model/endpoint listings on the Gemini Developer API pricing page checked on 2026-10-28, including Gemma 4's free-only family entry. Standard paid rates are primary; other listed paid tiers are recorded separately in strings. USD; not Vertex AI/Gemini Enterprise Agent Platform pricing. Tools and managed-agent service charges are outside these model tables. Status is checked against the official models directory and deprecation schedule; conflicted source claims are explicit.
All reference rows shown
Text and multimodal understanding
USD per 1M tokens; cache storage per 1M token-hours. Output includes thinking tokens.
| Model / status | Input | Output | Cache read | Cache storage | Other paid tiers | Scope & source |
|---|---|---|---|---|---|---|
| gemini-4-argonCurrent | $2 through 2026-11-30; $4 from 2026-12-01 | $10 through 2026-11-30; $20 from 2026-12-01 | $0.10 through 2026-11-30; $0.20 from 2026-12-01 | Not available: implicit caching only | Batch/Flex/Contributor input $1 then $2; output $5 then $10; cache $0.05 then $0.10. Priority input $3.60 then $7.20; output $18 then $36; cache $0.18 then $0.36. All changes effective 2026-12-01. | Paid-only standard access. Explicit cache unsupported. Contributor is invite-only private preview with data contribution; it is not equivalent to the ordinary paid data policy.Official source |
| gemini-3.8-flashCurrent | $0.75 through 2026-12-31; $1.50 from 2027-01-01 | $3.75 through 2026-12-31; $7.50 from 2027-01-01 | $0.075 through 2026-12-31; $0.15 from 2027-01-01 | $0.50 through 2026-12-31; $1 from 2027-01-01 | Batch/Flex/Contributor input $0.375 then $0.75; output $1.875 then $3.75; cache $0.0375 then $0.075. Priority input $1.35 then $2.70; output $6.75 then $13.50; cache $0.135 then $0.27. Storage follows Standard. Changes on 2027-01-01. | Stable. Contributor invite-only private preview uses submitted data for improvement. Free tier has limits.Official source |
| gemini-3.6-flashOlder generation | $0.75 through 2026-12-31; $1.50 from 2027-01-01 | $3.75 through 2026-12-31; $7.50 from 2027-01-01 | $0.075 through 2026-12-31; $0.15 from 2027-01-01 | $0.50 through 2026-12-31; $1 from 2027-01-01 | Batch/Flex input $0.375 then $0.75; output $1.875 then $3.75; cache $0.0375 then $0.075. Priority input $1.35 then $2.70; output $6.75 then $13.50; cache $0.135 then $0.27. Storage follows Standard. Changes on 2027-01-01. | Stable previous generation; no shutdown date announced.Official source |
| gemini-3.5-flash-liteCurrent | $0.30 text/image/video/audio | $2.50 | $0.03 | $1 | Batch/Flex input $0.15; output $1.25; cache $0.02. Priority input $0.54; output $4.50; cache $0.05. Contributor input $0.15; output $1.25; cache $0.015. Storage $1 in all listed tiers. | Standard all-input modalities share one token rate. Cache amounts are the published amounts, not inferred percentages. Contributor is invite-only private preview with data contribution.Official source |
| gemini-3.1-flash-liteOlder generation | $0.25 text/image/video; $0.50 audio | $1.50 | $0.025 text/image/video; $0.05 audio | $1 | Batch/Flex input $0.125 text/image/video, $0.25 audio; output $0.75; cache $0.0125 or $0.025 audio; storage $0.50. Priority input $0.45 or $0.90 audio; output $2.70; cache $0.045 or $0.09 audio; storage $1.80. | Stable older generation. Official earliest shutdown date: 2027-05-07; recommended successor gemini-3.5-flash-lite. Shutdown schedules can change.Official source |
| gemini-3.1-pro-previewPreview | $2 at prompt <=200k tokens; $4 above 200k | $12 at prompt <=200k tokens; $18 above 200k | $0.20 at prompt <=200k tokens; $0.40 above 200k | $4.50 | Batch/Flex input $1/$2 and output $6/$9 at <=200k/>200k; cache/storage unchanged. Priority input $3.60/$7.20; output $21.60/$32.40; cache $0.36/$0.72; storage $8.10. Contributor input $1/$2; output $6/$9; cache $0.10/$0.20; storage $4.50. | Paid-only. Long-prompt threshold changes output as well as input rate; no shutdown date announced. Contributor is invite-only private preview with data contribution.Official source |
| gemini-3.1-pro-preview-customtoolsPreview | $2 at prompt <=200k tokens; $4 above 200k | $12 at prompt <=200k tokens; $18 above 200k | $0.20 at prompt <=200k tokens; $0.40 above 200k | $4.50 | Batch/Flex input $1/$2 and output $6/$9 at <=200k/>200k; cache/storage unchanged. Priority input $3.60/$7.20; output $21.60/$32.40; cache $0.36/$0.72; storage $8.10. Contributor input $1/$2; output $6/$9; cache $0.10/$0.20; storage $4.50. | Paid-only. Long-prompt threshold changes output as well as input rate. Custom-tools endpoint uses the same published Pro pricing. Contributor is invite-only private preview with data contribution.Official source |
| gemini-3-flash-previewPreview | $0.50 text/image/video; $1 audio | $3 | $0.05 text/image/video; $0.10 audio | $1 | Batch/Flex input $0.25 text/image/video, $0.50 audio; output $1.50; cache/storage unchanged. Priority input $0.90 or $1.80 audio; output $5.40; cache $0.09 or $0.18 audio; storage $1.80. | Legacy preview; no shutdown date announced. Official replacement recommendation: gemini-3.6-flash.Official source |
| gemini-2.5-proLimited availability | $1.25 at prompt <=200k tokens; $2.50 above 200k | $10 at prompt <=200k tokens; $15 above 200k | $0.125 at prompt <=200k tokens; $0.25 above 200k | $4.50 | Batch/Flex input $0.625/$1.25; output $5/$7.50; cache/storage unchanged. Priority input $2.25/$4.50; output $18/$27; cache $0.225/$0.45; storage $8.10. Pairs mean <=200k/>200k. | Access restricted to previous active users. Official directory says these 2.5 text models are not deprecated; no shutdown announced.Official source |
| gemini-2.5-flashLimited availability | $0.30 text/image/video; $1 audio | $2.50 | $0.03 text/image/video; $0.10 audio | $1 | Batch/Flex input $0.15 text/image/video, $0.50 audio; output $1.25; cache/storage unchanged. Priority input $0.54 or $1.80 audio; output $4.50; cache $0.054 or $0.18 audio; storage $1.80. | Access restricted to previous active users; not deprecated; no shutdown announced.Official source |
| gemini-2.5-flash-liteLimited availability | $0.10 text/image/video; $0.30 audio | $0.40 | $0.01 text/image/video; $0.03 audio | $1 | Batch/Flex input $0.05 text/image/video, $0.15 audio; output $0.20; cache/storage unchanged. Priority input $0.18 or $0.54 audio; output $0.72; cache $0.018 or $0.054 audio; storage $1.80. | Access restricted to previous active users; not deprecated; no shutdown announced.Official source |
Audio, live conversation, transcription and speech
USD per 1M tokens by modality; approximate time equivalents are explicitly labeled and are not additional charges.
| Model / status | Input | Output | Other paid tiers | Scope & source |
|---|---|---|---|---|
| gemini-3.8-liveCurrent | $0.75 text; $3 audio (published approx. $0.005/min); $1 image/video (published approx. $0.002/min) | $4.50 text; $12 audio (approx. $0.018/min) | Standard pricing shown; no additional paid tier table | Stable Live API. The extended-thinking endpoint has the same published rates. Actual consumption, including thinking where applicable, determines the bill.Official source |
| gemini-3.8-live-extended-thinkingCurrent | $0.75 text; $3 audio (published approx. $0.005/min); $1 image/video (published approx. $0.002/min) | $4.50 text; $12 audio (approx. $0.018/min) | Standard pricing shown; no additional paid tier table | Stable Live API. The extended-thinking endpoint has the same published rates. Actual consumption, including thinking where applicable, determines the bill.Official source |
| gemini-3.1-flash-live-previewDeprecated | $0.75 text; $3 audio (published approx. $0.005/min); $1 image/video (published approx. $0.002/min) | $4.50 text; $12 audio (approx. $0.018/min) | Standard pricing shown; no additional paid tier table | Legacy preview. Earliest shutdown 2026-11-17; migrate to gemini-3.8-live.Official source |
| gemini-3.5-live-translate-previewPreview | $3.50 audio (approx. $0.0053/input minute) | $21 audio (approx. $0.0315/output minute) | Standard only on pricing page | Input and output are separately metered at 25 audio tokens/second. Published combined equivalent is approximately $0.0368/min when input and output durations match; no shutdown announced.Official source |
| gemini-3.5-transcribe-liveCurrent | $3.50 audio (published approx. $0.005/min) | $21 text (published approx. $0.004/min) | Standard only on pricing page | Time estimates assume 25 audio tokens/second input and 175 text tokens/minute output; published combined estimate about $0.009/min; no search grounding.Official source |
| gemini-3.5-transcribeCurrent | $2 audio (approx. $0.003/min) | $12 text (published approx. $0.002/min) | Standard only; model documentation says Batch/Flex/Priority unsupported | Time estimate assumes 25 audio tokens/second input and 175 text tokens/minute output; published combined estimate about $0.005/min; no search grounding.Official source |
| gemini-3.8-flash-ttsCurrent | $0.50 text through 2026-12-31; $1 from 2027-01-01 | $9 audio through 2026-12-31; $18 from 2027-01-01. Equivalent $0.00225 then $0.0045 per 10s audio. | Batch/Flex input $0.25 then $0.50; audio output $4.50 then $9 (equivalent $0.001125 then $0.00225/10s). Priority input $0.90 then $1.80; output $16.20 then $32.40 (equivalent $0.00405 then $0.0081/10s). All changes 2027-01-01. | Stable TTS; audio uses 25 tokens/second. Time equivalent covers audio output only; text input is additional.Official source |
| gemini-3.8-flash-lite-ttsCurrent | $0.50 text through 2026-12-31; $1 from 2027-01-01 | $6 audio through 2026-12-31; $12 from 2027-01-01. Equivalent $0.0015 then $0.003 per 10s audio. | Batch/Flex input $0.25 then $0.50; output $3 then $6 (equivalent $0.00075 then $0.0015/10s). Priority input $0.90 then $1.80; output $10.80 then $21.60 (equivalent $0.0027 then $0.0054/10s). All changes 2027-01-01. | Stable TTS; audio uses 25 tokens/second. Time equivalent covers audio output only.Official source |
| gemini-3.1-flash-tts-previewDeprecated | $1 text | $20 audio | Batch input $0.50 text; output $10 audio | Legacy preview. Earliest shutdown 2026-11-17; replacements gemini-3.8-flash-tts or gemini-3.8-flash-lite-tts. 25 audio tokens/second.Official source |
| gemini-2.5-flash-native-audio-preview-12-2025Deprecated | $0.50 text; $3 audio/video | $2 text; $12 audio | No additional paid tier table | Older Live API preview. Earliest shutdown 2026-11-17; replacement gemini-3.8-live. Existing-user restrictions may also apply.Official source |
| gemini-2.5-flash-preview-ttsDeprecated | $0.50 text | $10 audio | Batch input $0.25 text; output $5 audio | Older preview. Earliest shutdown 2026-11-17; replacements gemini-3.8-flash-tts or gemini-3.8-flash-lite-tts. Existing-user restrictions may also apply.Official source |
| gemini-2.5-pro-preview-ttsDeprecated | $1 text | $20 audio | Batch input $0.50 text; output $10 audio | Older preview. Earliest shutdown 2026-11-17; replacements gemini-3.8-flash-tts or gemini-3.8-flash-lite-tts. Paid-only access; existing-user restrictions may also apply.Official source |
Image generation and editing
USD per 1M tokens unless a value explicitly says per image. Image equivalents are alternatives, not extra charges.
| Model / status | Input | Text output | Image output | Other paid tiers | Scope & source |
|---|---|---|---|---|---|
| gemini-nano-banana-2.1Current | $1.50 text/image/video | $7.50 text/thinking | $30 image tokens; approximately $0.0336/1K image, $0.0504/2K, $0.113/4K | Batch input $0.75; text output $3.75; image tokens $15, equivalent $0.0168/1K, $0.0252/2K, $0.0567/4K. | Paid-only. Output image token counts: 1120 at 1K, 1680 at 2K, 3780 at 4K. Preserve vendor rounding in equivalents.Official source |
| gemini-3.1-flash-imageCurrent | $0.50 text/image | $3 text/thinking | $60 image tokens; approx. $0.045/0.5K image, $0.067/1K, $0.101/2K, $0.151/4K | Batch input $0.25; text output $1.50; image tokens $30; approx. $0.022/0.5K, $0.034/1K, $0.050/2K, $0.076/4K. | Nano Banana 2. Paid-only. Output image token counts: 747/1120/1680/2520 for 0.5K/1K/2K/4K; no shutdown announced.Official source |
| gemini-3.1-flash-lite-imageCurrent | $0.25 text/image/video | $1.50 text/thinking | $30 image tokens; $0.0336 per 1K image | Batch input $0.125; text output $0.75; image tokens $15; $0.0168/1K image. | Nano Banana 2 Lite. Paid-only. A 1K output image uses 1120 tokens.Official source |
| gemini-3-pro-imageCurrent | $2 text/image; approx. $0.0011 per input image | $12 text/thinking | $120 image tokens; approx. $0.134/1K or 2K image; $0.24/4K | Batch/Flex input $1 per 1M text tokens or approx. $0.0006/input image; text output $6; image output $0.067/1K or 2K, $0.12/4K. Priority input $3.60; text output $21.60; image-token output $216. | Nano Banana Pro. Paid-only. Input image uses 560 tokens; output uses 1120 tokens for 1K/2K or 2000 for 4K. Per-image and token charges are alternate expressions.Official source |
| gemini-2.5-flash-imageDeprecated | $0.30 text/image | Page points to 2.5 Flash text pricing: $2.50 | $0.039 per image; footnote $30 per 1M image tokens | Batch/Flex input $0.15; output $0.0195/image. Priority input $0.54; output $0.0702/image. | Availability conflict: pricing warning says shutdown 2026-10-02, but deprecations table lists 2027-03-15. Published rates retained for reference; verify endpoint availability before use. Image equivalent assumes 1290 tokens up to 1024x1024. Do not treat this row as a recommended new deployment.Official source |
Video generation and editing
USD per 1M tokens; video time equivalent is approximate and not an extra charge.
| Model / status | Input | Text output | Video output | Scope & source |
|---|---|---|---|---|
| gemini-omni-1.1-flashCurrent | $1.50 text/image/video/audio | $9 | $17.50 video tokens; approximately $0.10 per second of 720p video | Paid-only GA endpoint. Published conversion uses 5792 output tokens per second at 720p; other resolutions are not assigned an invented time rate; no other paid tier table is listed.Official source |
Music generation
USD per generated song/request, not per token or minute.
| Model / status | Generation price | Length scope | Scope & source |
|---|---|---|---|
| lyria-3.5Current | $0.08/song | Full song | Paid-only; no per-minute rate is inferred.Official source |
| lyria-3-clip-previewPreview | $0.04/song | 30-second clip | Legacy preview family; no shutdown date announced.Official source |
| lyria-3-pro-previewPreview | $0.08/song | Full song | Legacy preview; successor lyria-3.5; no shutdown date announced.Official source |
Embeddings
USD per 1M input tokens by modality; per-item/time equivalents are alternatives.
| Model / status | Text input | Image input | Audio input | Video input | Batch | Scope & source |
|---|---|---|---|---|---|---|
| gemini-embedding-2Current | $0.20 | $0.45; approx. $0.00012/image | $6.50; approx. $0.00016/second | $12; approx. $0.00079/frame | Text $0.10; image $0.225 (~$0.00006/image); audio $3.25 (~$0.00008/second); video $6 (~$0.000395/frame). | Stable model card confirms gemini-embedding-2 despite an older directory entry still using a preview name. PDFs/DOCUMENT modality follow image-token rates; no output-token charge is listed.Official source |
Robotics and embodied reasoning
USD per 1M tokens; cache storage per 1M token-hours.
| Model / status | Input | Output | Cache read | Cache storage | Other paid tiers | Scope & source |
|---|---|---|---|---|---|---|
| gemini-robotics-er-2-previewPreview | $1 text/image/video/audio through 2026-12-31; $2 from 2027-01-01 | $5 through 2026-12-31; $10 from 2027-01-01 | $0.10 through 2026-12-31; $0.20 from 2027-01-01 | $0.50 through 2026-12-31; $1 from 2027-01-01 | Batch input $0.50 then $1; output $2.50 then $5; cache $0.05 then $0.10; storage same as Standard. Changes 2027-01-01. | Embodied reasoning preview; no shutdown announced. Output includes thinking tokens.Official source |
| gemini-robotics-er-2-streaming-previewPreview | $1 text/image/video/audio through 2026-12-31; $2 from 2027-01-01 | $5 through 2026-12-31; $10 from 2027-01-01 | Not listed | Not listed | Standard only on pricing page | Live API text-streaming robotics endpoint; no shutdown announced. Do not copy non-streaming cache or Batch rates into this endpoint.Official source |
Free-only model-family listing
Gemini Developer API availability; no paid USD token rate is offered here.
| Model / status | Input | Output | Cache read | Cache storage | Scope & source |
|---|---|---|---|---|---|
| Gemma 4Current | Free tier: $0; paid tier unavailable | Free tier: $0; paid tier unavailable | Free tier: $0; paid tier unavailable | Free tier: $0; paid tier unavailable | The pricing page lists the family, not individual size endpoints. Free-tier limits and data policy apply. Do not represent unavailable paid capacity as a $0 paid offer or infer infrastructure cost for self-hosting.Official source |
A retained price row does not establish current API availability. Older, deprecated and retired models are labelled separately. Units, prompt tiers and access conditions are part of each price; consult the linked source before budgeting.
Calculator preset billing notes
The calculator uses the four selected text-model presets shown in its menu. The full reference above also includes models with other units, access conditions or retired status.
Higher prompt tiers
| Model | Threshold per request | Input / 1M | Output / 1M |
|---|---|---|---|
| Gemini 3.1 Pro Preview | Over 200,000 input tokens | $4 | $18 |
Paid API use versus consumer plans
These figures cover selected paid Gemini Developer API text models. Consumer Gemini subscriptions, Vertex AI billing and free-tier eligibility are separate. Use the pricing source for your actual account and route.
Announced rate changes
The source snapshot includes announced increases for Gemini 4 Argon on 1 December 2026 and Gemini 3.8 Flash on 1 January 2027. With JavaScript enabled, the calculator applies recorded rate changes by the current date. The reference tables show the checked rates alongside announced effective dates. New or unannounced changes require a source check and rebuild.
Thinking, caching and long context
Output counts must include billed thinking tokens. Explicit cache storage can be billed separately. The selected Gemini 3.1 Pro Preview record changes rate above 200,000 input tokens per request. Preview status and usage limits should be checked before implementation.
Cache-write reference: base prompt tier
- Gemini 4 Argon: No separately verified cache-write rate in this record.
- Gemini 3.8 Flash: No separately verified cache-write rate in this record.
- Gemini 3.5 Flash-Lite: No separately verified cache-write rate in this record.
- Gemini 3.1 Pro Preview: No separately verified cache-write rate in this record.
The cached-input share models reads. Newly written tokens remain in ordinary input usage. To avoid counting their input charge twice, add only the write premium above that already-counted input charge, plus full storage costs, under other monthly costs. For example, 1M Sonnet input tokens at $2 with a $2.50 write rate need a $0.50 adjustment, not another $2.50. These reference rates use the base prompt tier; check the source for long-context write rates.