Gemini 3.7 Flash is Google's August 13, 2026 coding and agent workhorse. Introductory API price is $0.75 per million input tokens and $3.75 per million output tokens through 31 December 2026, then $1.50 and $7.50. Context window is 1,048,576 tokens. o10 lists it at band C2 on Google Gemini and OpenRouter.
Last updated: 2026-08-14. Full catalog. Reference prices are dated snapshots. Compare total cost and measured task quality before choosing a route.
Gemini 3.7 Flash is Google's 13 August 2026 Flash-series workhorse for coding, agents, and knowledge work. It ships three weeks after Gemini 3.6 Flash. Google positions it as Pro-level agentic capability at Flash economics, with text, image, audio, and video input and text output. Model ID: gemini-3.7-flash.
Google blog, 13 August 2026.
Introductory global price through 31 December 2026 is $0.75 per million input tokens and $3.75 per million output tokens, with context cache at $0.075 per million. On 1 January 2027 that doubles to $1.50 input and $7.50 output. Google Cloud also publishes regional and batch variants; use the venue you actually call.
Google Cloud Agent Platform pricing.
Gemini 3.7 Flash has a 1,048,576-token context window and a 65,536-token maximum output. That is the 1M Flash envelope, not a 128K chat model. Host-specific gateways may still cap lower; check the endpoint.
Google Cloud model card, August 2026.
Developers get Gemini 3.7 Flash via the Gemini API in Google AI Studio and Android Studio, plus Google Antigravity. Enterprises get Gemini Enterprise Agent Platform. Google AI Pro/Ultra subscribers get it through Spark in supported countries. o10 maps gemini-3-7-flash to Google Gemini and OpenRouter.
o10 fulfillment overlay 2026-08-14.
No. Google shipped 3.7 Flash before a new Gemini 3.5 Pro GA. Reuters and Axios noted 3.5 Pro was still in partner testing when 3.7 Flash launched. Treat 3.7 Flash as the cost-efficient agent workhorse, not the flagship Pro. Keep a Pro/C3 candidate on the eval ladder.
Google launch coverage, August 2026.
Gemini 3.7 Flash is band C2. At $0.75/$3.75 intro pricing it is cheap enough to be a default for many coding and agent evals, and expensive enough that you still want a C0/C1 fallback when the floor is lower. o10 will not pin it unless evals require it.
o10 band C2.
Gemini 3.7 Flash in the o10 unified catalog.
| Field | Value |
|---|---|
| o10 slug | `gemini-3-7-flash` |
| Type | text |
| Band | C2 |
| Routable | Yes |
| Source provider | |
| Primary host | Google Gemini |
| Gateway ID | `google/gemini-3.7-flash` |
| Context window | 1,048,576 tokens |
| Released | 2026-08-13 |
Gateway snapshot pricing for Gemini 3.7 Flash.
| Unit | Price |
|---|---|
| Input | $0.75/1M tokens |
| Output | $3.75/1M tokens |
Upstream venues configured for Gemini 3.7 Flash.
| Host | Upstream model ID |
|---|---|
| gemini | gemini/gemini-3.7-flash |
| openrouter | openrouter/google/gemini-3.7-flash |
Eval-gated model selection for this slug.
o10 slug: gemini-3-7-flash. Band C2 (workhorse). Routable. Strong default candidate for coding and agent jobs when evals pass; o10/auto can prefer it over C3/C4 when the floor is met.
See [/models/routing](/models/routing) for smart aliases (o10/auto, o10/frontier, o10/squad) and [/kyi](/kyi) for governance.
Introductory rates are temporary. Check their end date before forecasting costs.
Google priced 3.7 Flash at half of original 3.6 Flash through 31 December 2026. That is a temporary spread, not a permanent Flash price. Teams that bake $0.75/$3.75 into 2027 budgets will miss the 1 January step-up to $1.50/$7.50.
o10 shadow receipts should stamp both the intro rate and the scheduled standard rate so finance sees the cliff before enforce.
Specs from Google Cloud Gemini 3.7 Flash docs (model ID gemini-3.7-flash, 1,048,576 context, 65,536 max output). Introductory prices from Google's 13 August 2026 announcement; standard prices apply 1 January 2027. o10 overlay snapshot 2026-08-14.
Intro price through 31 December 2026: $0.75/1M input and $3.75/1M output. From 1 January 2027: $1.50 and $7.50. Cache is $0.075 intro, $0.15 standard.
1,048,576 input tokens and 65,536 max output tokens per Google Cloud docs for gemini-3.7-flash.
13 August 2026, about three weeks after Gemini 3.6 Flash. GA model ID gemini-3.7-flash.
Yes. Modalities are text in/out plus image, audio, and video input. Output is text (including reasoning tokens billed as output).
Send model "gemini-3-7-flash" or gateway id "google/gemini-3.7-flash". o10/auto can select it when it is the cheapest eval-passing C2 candidate.
Band C2, eval-gated. After 1 January 2027 the doubled list price can change whether auto still prefers it versus C1 open-weight or Grok 4.6. Re-run shadow when the intro window ends.
o10 slug "gemini-3-7-flash"; Google id "gemini-3.7-flash"; gateway "google/gemini-3.7-flash". Snapshot 2026-08-14.