GEMINI 3.6 GA1M CONTEXTMULTIMODALCOMPUTER USEHALF PRICE UNTIL 2027
Gemini 3.6 Flash API Pricing
Gemini 3.6 Flash shipped GA on July 21, 2026 and is now Google's previous-generation Flash model, superseded by Gemini 3.7 Flash. Google has put it on a dated schedule: $0.75/M input, $3.75/M output and $0.075/M cached input through December 31, 2026, then standard pricing of $1.50 / $7.50 / $0.15 from January 1, 2027. The rates below are the ones in effect today — budget the January doubling before it lands. Checked by hand against ai.google.dev.
Input - per 1M tokens
$0.75/M
$1.50 from Jan 1, 2027 -50%
Output - per 1M tokens
$3.75/M
$7.50 from Jan 1, 2027 -50%
Cached input - per 1M tokens
$0.075/M
$0.15 from Jan 1, 2027 -90%
Effective - agentic blend
$0.481/M
92/8 split - 82% cache
§ 01 / TERMINAL
Gemini 3.6 Flash API cost calculator.
Live calculator pre-loaded with the Gemini 3.6 Flash rates in effect today — $0.75 input, $3.75 output, $0.075 cached. On January 1, 2027 every one of those doubles, so a workload sized here will cost twice as much then. Tweak spend, output mix, or cache assumptions and share the URL to share the calculation.
$ /mo
Workload split
Prompt cache hit rate
Tokens you can process
—
Words equivalent (English)
—
Effective rate
—
§ 02 / SCENARIOS
Gemini 3.6 Flash cost on real workloads.
CODING AGENT
Codebase-scale migration
$0.061/task
LONG DOC
Reading 100-page contracts
$0.065/doc
RAG SUPPORT
Ticket triage
$0.00532/ticket
ASSISTANT
Research planning turn
$0.011/turn
Estimate · gemini-tokenizer-estimate · ≈3.85 chars/token Auto-counts as you type
This is a chars-per-token approximation, not a real tokenizer. Actual tokens vary by language, code density, and tool-call overhead — counts are typically ±10–20% off for English prose, more for code or non-Latin scripts. For exact billing, use the vendor's official tokenizer.
Characters 674
Words 109
Tokens (estimated) 175 tokens
Cost as input · uncached $0.00013 USD
Cost as output · uncached $0.00066 USD
Cost as cached input $0.00001 USD
| Model | Input /M | Output /M | Effective blended | Context | Best for |
|---|---|---|---|---|---|
| Gemini 3.6 Flash Current | $0.75 cache $0.075 | $3.75 | $0.481 agentic 92/8 | 1M | Mainstream agentic + multimodal work |
| Gemini 3.5 Flash | $1.50 cache $0.15 | $9.00 | $1.08 deprecated · +2.2× | 1M | Deprecated; its calls now route here |
| Gemini 3.1 Pro Preview | $2.00 cache $0.20 | $12.00 | $1.44 +3.0× | 1M | Google's frontier reasoning tier |
| Gemini 3.5 Flash-Lite | $0.30 cache $0.03 | $2.50 | $0.272 −43% | 1M | High-volume, simpler tasks |
| Gemini 3.1 Flash-Lite | $0.25 cache $0.025 | $1.50 | $0.18 deprecated · −2.7× | 1M | Cheaper Lite tier, retiring May 2027 |
| Claude Haiku 5.5 | $0.10 cache $0.01 | $0.50 | $0.0682 −7.0× | 1M | Anthropic's fast, high-volume tier |
| GPT-6.1 Terra | $1.00 cache $0.05 | $5.00 | $0.645 +34% | 1.05M | OpenAI's fast, efficient GPT-6 tier |
| DeepSeek V4.1 Flash | $0.30 cache $0.006 | $1.20 | $0.15 −3.2× | 1M | Budget reasoning and coding |
Gemini 3.6 Flash has published scores on 14 benchmarks we track, and leads 2 of them. 2 of these scores come from an independent leaderboard; the rest are the lab's own reports.
| Benchmark | Score | Rank | Leader | Source |
|---|---|---|---|---|
| SWE-Bench Pro Coding · Public set (731 tasks) | 58.7% public set | #1 of 6 | Gemini 3.6 Flash leads | Vendor-reported Google DeepMind |
| MLE-bench AI research | 63.9% as reported | #1 of 5 | Gemini 3.6 Flash leads | Vendor-reported Google DeepMind |
| OSWorld-Verified Agents and computer use | 83% as reported | #2 of 8 | Kimi K3 · 84.8% | Vendor-reported Google DeepMind |
| ARC-AGI-1 Reasoning | 91.2% ARC Prize verified, best result | #28 of 99 | 6 tied · 98.5% | Official leaderboard ARC Prize Foundation |
| MRCR v2 Long context · 8-needle, up to 128K | 91.8% 128K average | #2 of 7 | Gemini 3.7 Flash · 97% | Vendor-reported Google DeepMind |
| ARC-AGI-2 Reasoning | 60.4% ARC Prize verified, best result | #35 of 105 | GPT-6 Astra · 95% | Official leaderboard ARC Prize Foundation |
| CharXiv Reasoning Vision and documents · No tools | 85.2% no tools | #3 of 9 | Kimi K3.1 Pro · 89% | Vendor-reported Google DeepMind |
| Terminal-Bench 2 Agents and computer use · Terminal-Bench 2.1 | 78% as reported | #12 of 17 | DeepSeek V4.1 Flash · 90.6% | Vendor-reported Google DeepMind |
| Agents' Last Exam Agents and computer use · Strict pass rate | 24.2% pass rate | #4 of 5 | Gemini 4 Argon · 39.5% | Vendor-reported Google DeepMind |
| DeepSWE Coding | 48.6% as reported | #23 of 27 | Gemini 4 Argon · 77.9% | Vendor-reported Google DeepMind |
| Terminal-Bench 3.0 Agents and computer use | 5.4% as reported | #8 of 9 | DeepSeek V4.1 Flash · 30% | Vendor-reported Google DeepMind |
| AutomationBench Agents and computer use · Private set | 17% private set | #2 of 2 | Gemini 3.7 Flash · 30.4% | Vendor-reported Google DeepMind |
| FrontierCode Coding · v1.1 Main (100 hardest tasks) | 34.4% score | #13 of 13 | Claude Opus 5.5 · 54.4% | Vendor-reported Google DeepMind |
| GDP.pdf Vision and documents · Default setting | 22% as reported | #7 of 7 | Gemini 3.8 Flash · 35% | Vendor-reported Google DeepMind |
Gemini 3.6 Flash API pricing FAQ.
Practical pricing questions, separated from calculator assumptions and regional taxes.
Q · 01 What is Gemini 3.6 Flash priced at? +
Google's paid-tier pricing page lists
$0.75/M input, $3.75/M output and $0.075/M cached input through December 31, 2026; standard pricing of $1.50 / $7.50 / $0.15 starts January 1, 2027. Explicit cache storage is billed on top at $0.50 per 1M tokens per hour, rising to $1.00 on the same date.Q · 02 Is it cheaper than Gemini 3.5 Flash? +
Yes, and by far more than at launch. It shipped 17% cheaper on output than 3.5 Flash at identical input; since August 14, 2026 the whole card is halved to
$0.75 / $3.75 against 3.5 Flash's unchanged $1.50 / $9.00. On our agentic blend that is $0.48/M effective against $1.08/M for 3.5 Flash — less than half. The halved rate is scheduled to expire on December 31, 2026. Google also says 3.6 consumes fewer output tokens per task, which would compound the saving, but that is a vendor claim we have not measured.Q · 03 Does output pricing include thinking tokens? +
Yes. Google's price table labels the output row "including thinking tokens", so reasoning is billed at the same
$3.75/M. There is no separate reasoning rate and no cheaper non-thinking variant.Q · 04 How is the effective price calculated? +
AI//COST uses the same 92/8 agentic blend everywhere: 92% of tokens are input, 82% of which are served from cache. For Gemini 3.6 Flash that gives a blended input of
$0.393/M and an effective cost of $0.96/M. Every model in the shelf above uses the identical formula, so the numbers compare directly.Q · 05 What does Batch or Flex inference cost? +
Batch and Flex are half the standard rate:
$0.375/M input and $1.88/M output, with cached input at half the standard cache rate too. Priority costs 1.8x the standard rate on every line. Like the standard tier, all three double on January 1, 2027. Read from Google's pricing page on October 13, 2026.Q · 06 Is Search grounding included in the token price? +
No. Grounding with Google Search and with Google Maps is billed separately: 5,000 prompts per month free, shared across the Gemini 3 family, then
$14 per 1,000 queries. One request can trigger several queries, and each is charged.Q · 07 What context window does it support? +
1,048,576 input tokens and 65,536 output tokens. Inputs cover text, image, video, audio and PDF; output is text only. Computer use is supported in preview, while Live API, audio generation and image generation are not.
Q · 08 How accurate is the tokenizer estimate? +
The browser widget uses a
gemini-tokenizer-estimate chars-per-token ratio for English text — Google publishes no downloadable tokenizer, so this is an estimate, not exact counting. Actual billing comes from the API's usage metadata and differs for code, non-Latin scripts and multimodal input.