If you’ve been following Google’s aggressive AI shipping tempo, you already know the Flash lineup is where the real speed lives. But a new wave of developer sleuthing suggests the next iteration—Gemini 3.7 Flash—might be much closer to launch than anyone anticipated.
The buzz erupted after references to Gemini 3.7 Flash surfaced directly in Google’s official Python GenAI SDK codebase. As reporting from NokiaPowerUser highlights, Google appears to already be actively laying the groundwork for this next-gen lightweight model.
Is a release imminent, or is this just standard developer staging? Here is a clear breakdown of what leaked, what looks credible, and what you should take with a pinch of salt.
1. The Concrete Signal: Python GenAI SDK Code Drops
Unlike typical rumors sparked by anonymous social media posts, this signal has developer receipts.
Engineers spotted explicit references to Gemini 3.7 Flash inside Google’s public google-genai Python SDK repository. In API development, when a model string lands in active client libraries, it usually means internal testing and staging are nearing completion.
Why it matters: Google updates its Flash models at a rapid pace. With the Flash family evolving so fast, a quick follow-up to previous iterations fits Google’s bi-monthly cadence for lightweight, high-speed models.
2. The Pricing Rumor: $0.75 / $3.75 per Million Tokens
Along with the SDK discovery, community reports started floating potential pricing specs for Gemini 3.7 Flash:
Input Tokens: $0.75 per 1,000,000 tokens
Output Tokens: $3.75 per 1,000,000 tokens
If those figures hold up, it would cut operational costs drastically compared to prior Flash pricing, turning 3.7 Flash into a budget-friendly powerhouse for agentic workflows and high-volume applications.
The reality check: Google hasn’t confirmed these rates anywhere in its official Google AI Studio or Vertex AI documentation. Treat these price points as preliminary figures from test configurations until an official model card drops.
3. Where Does This Leave Gemini 3.5 Pro and Gemini 4?
While the Flash line moves full speed ahead, the flagship Pro roadmap looks a bit more complicated:
Gemini 3.5 Pro Delays: Developers expecting a mid-generation flagship refresh have been left waiting after quiet timeline shifts.
The “3.5 Pro is Dead” Whispers: Rumors on tech X and forum boards suggest Google might bypass 3.5 Pro altogether to put full momentum into Gemini 4.
What’s actually true: Google hasn’t officially canceled 3.5 Pro. Tech companies often skip minor point updates internally to align major releases, but right now, claims that 3.5 Pro is officially dead remain unconfirmed spec.
4. Measuring the Launch Odds
Looking at current insider forecasts (including analysis like Vepsi’s Bet), here is how the odds stack up:
[==================== 68% ====================] Chance of Launch "Soon" (Days to Weeks)
[====== 28% ======] Chance of Dropping "Later Today"
68% Chance of a near-term release: Highly plausible. Google rarely commits new model identifiers to public SDK branches unless production deployment is imminent.
28% Chance of a same-day drop: Unlikely. While code merges often precede launches, exact launch-day leaks are historically noisy and prone to shifting release windows.
Fact vs. Rumor: The Quick Breakdown
| Topic | Status | What We Know |
| SDK Code Leak | Verified Signal | Real: Model hooks exist in the official Python SDK. |
| $0.75 / $3.75 Pricing | Unconfirmed | Unverified: Sound appealing, but lack official documentation. |
| Release Timeline | High Likelihood | Imminent: Launching soon, though a same-day drop is a long shot. |
| 3.5 Pro Cancellation | Pure Speculation | Unproven: Google has made no official announcement on sunsetting 3.5 Pro. |
What Developers Should Watch For Next
The SDK reference confirms that Google is actively prepping Gemini 3.7 Flash behind the scenes. If you’re building on the Gemini ecosystem, keep a close eye on Google AI Studio announcements over the coming days—just hold off on updating your application budget until Google publishes the official pricing tier.
