Google announces Gemini 3.7 Flash with introductory API pricing
Google’s 13 August 2026 announcement says Gemini 3.7 Flash is available through specified developer and enterprise routes, with introductory API prices of $0.75 per million input tokens and $3.75 per million output tokens through the end of 2026. Google also reports improved results over Gemini 3.6 Flash on five named evaluations and says the release includes updated safeguards. The supplied evidence does not independently validate those results or establish availability, pricing eligibility, or terms for Armenian accounts. [1]
This edition passed Imananq's enhanced publication checks. Some material claims remain explicitly attributed to official or company sources because no independent source is currently bound to this edition. The engine continues checking approved sources and will add corroboration only through a new edition that passes the full gate.
Google announced Gemini 3.7 Flash on 13 August 2026, describing it as a model for coding and agent workflows. The company says developers can access it through Antigravity, the Gemini API via Google AI Studio, and Android Studio, while enterprise access is offered through Gemini Enterprise products. [1]
01
What we know now
- 01
[1] Google, “Introducing Gemini 3.7 Flash,” published 13 August 2026: https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-3-7-flash (complete primary evidence).
- 02
[1] The source is a first-party product announcement. Its capability, benchmark, pricing, and safety statements are attributed to Google.
02
Announcement published
Google published the Gemini 3.7 Flash announcement on 13 August 2026.Introductory input price
Google states an introductory API price of $0.75 per million input tokens through the end of 2026; the precise closing time is not stated.Introductory output price
Google states an introductory API price of $3.75 per million output tokens through the end of 2026; availability by account and country is not established here.Reported coding benchmark
Google reports 43.6% for Gemini 3.7 Flash and 34.4% for Gemini 3.6 Flash on FrontierCode 1.1 Main. The supplied evidence does not include methodology or independent validation.All performance and availability details shown are from Google’s first-party announcement and are not independently verified in the supplied record.
03
What Google says is available
Google’s first-party announcement establishes that it is introducing Gemini 3.7 Flash and identifies several product routes through which it says the model can be used. It does not provide supported-country, account-eligibility, rate-limit, or restriction details for those routes in the supplied material. [1]
- Google says the release follows Gemini 3.6 Flash by three weeks.
- For developers, Google lists Google Antigravity, the Gemini API through Google AI Studio, and Android Studio.
- For enterprises, Google lists Gemini Enterprise Agent Platform and the Gemini Enterprise app.
- Google also says its Gemini Spark product will begin using the model for eligible Google AI Pro and Ultra subscribers in supported countries.
04
Introductory API price
Google says Gemini 3.7 Flash is offered at introductory API prices of $0.75 per million input tokens and $3.75 per million output tokens through the end of 2026. The announcement does not specify the final date, closing time, or time zone, and it does not confirm that the prices apply to accounts in Armenia. [1]
- Input tokens: $0.75 per 1 million tokens.
- Output tokens: $3.75 per 1 million tokens.
- Stated period: through the end of 2026.
05
Reported benchmark comparisons
Google reports higher results for Gemini 3.7 Flash than Gemini 3.6 Flash across the named coding, web-development, document-processing, and automation evaluations. These are Google’s own reported comparisons, not independently established findings: the supplied announcement does not include the underlying test conditions, evaluation methodology, or reproducibility material. [1]
- FrontierCode 1.1 Main: 43.6% for 3.7 Flash versus 34.4% for 3.6 Flash.
- DeepSWE v1.1: 65.3% versus 49.0%.
- Arena.ai WebDev Arena: 1588 Elo versus 1538.
- GDP.pdf: 34.0% versus 22.0%.
- AutomationBench: 30.4% versus 17.0%.
06
Safety update: what is confirmed and what is not
Google says Gemini 3.7 Flash ships with updated Frontier Safety safeguards in the cited areas. That confirms the company’s stated product position, but the supplied evidence does not allow an assessment of safeguard effectiveness, testing, limitations, or the content of the referenced model card. [1]
- Google says the model includes updated safeguards relating to chemical, biological, radiological and nuclear misuse and cyber offense.
- The announcement references a Gemini 3.7 Flash model card, but that document was not included in the evidence packet.
07
What to check before adopting it
Google’s announcement provides an access and price claim, but does not settle account eligibility or operating terms for every location. [1]
- 01
Confirm that Gemini API, Google AI Studio, Android Studio, Antigravity, and any enterprise product are available for the intended account and use case. [1]
- 02
Verify current API pricing, rate limits, supported-country status, and the precise expiry of introductory pricing in the applicable product terms before committing a budget. [1]
- 03
Test relevant workflows against an existing model and review the separately linked Gemini 3.7 Flash model card; its contents were not included in the supplied evidence. [1]
08
Limits of this edition
The supplied record contains one first-party Google announcement and no independent reporting or evaluation replication. [1]
Google’s benchmark comparisons do not come with methodology, reproducibility materials, or independent validation in the supplied evidence. [1]
The announcement does not establish availability, pricing eligibility, rate limits, or usage restrictions for accounts in Armenia. [1]
Google says introductory pricing runs through the end of 2026, but does not give an exact closing date, time, or time zone. [1]
A model card is linked by Google but was not supplied as evidence, so its safety findings, testing details, and limitations cannot be assessed here. [1]
SRC
Source desk
Direct links to the material behind this selection. Seeing the source matters as much as reading the synthesis.
Keep exploring
18 Aug 2026
AI agents: from promise to an accountable workflow
A plain-language anatomy of agents, their failure modes, and the controls that make delegated work observable and reversible.
25 Aug 2026
Qwen documents Qwen3.8-2.4T-A95B checkpoint, with text-only and mandatory-thinking limits
Qwen’s repository documents a very large text-generation checkpoint in Transformers format and gives deployment pointers for vLLM, SGLang, and TokenSpeed. Its most consequential documented constraints are text-only input and unavoidable thinking output, while licensing terms, infrastructure needs, and independent performance validation remain unknown in the supplied record. [1]
21 Aug 2026
Qwen checkpoint listing: registry facts and remaining documentation gaps
The supplied registry record confirms a Qwen-listed repository with image-text-to-text, transformers, and apache-2.0 tags. It does not supply the model-card evidence needed to evaluate capabilities, requirements, access conditions, intended use, or full license terms.