Gemini 3.5 Pro: When Does It Release?
As of 2026-10-07, no release date for Gemini 3.5 Pro turns up on the official Gemini API models page or release notes: the newest Pro model listed there is Gemini 3.1 Pro (gemini-3.1-pro-preview). The most recent official wording, from the 2026-07-21 blog, is only that it is "currently testing with partners" and will be made "broadly available as soon as it's ready" — with no date. The same post announced that Google had begun its most ambitious pre-training run yet, for Gemini 4. What actually shipped on 09-02 was Gemini 3.8 Flash. Here is the verified timeline and what is confirmed versus rumoured.
One key across Claude, GPT and Chinese models, billed per token.
- Billed per token — live rates on /models
- Pay by card (Visa / Mastercard / AMEX), Apple Pay, Google Pay or crypto
- Self-service: top up and activate instantly
As of 2026-10-07, no record of a Gemini 3.5 Pro release turns up on the official Gemini API models page or release notes. The newest Pro model on the official Gemini API models page is Gemini 3.1 Pro (gemini-3.1-pro-preview), with no 3.5 Pro entry, and the newest entry in the official Gemini API release notes, dated 2026-10-06 (Gemini Nano Banana 2.1 generally available), does not mention 3.5 Pro either. On 2026-09-18, while limiting access to the 2.5 models, the release notes told new projects to use 3.5 Flash-Lite or 3.8 Flash. The most recent checkable official statement is from the 2026-07-21 blog: "Gemini 3.5 Pro is currently testing with partners and we plan to make it broadly available as soon as it's ready" — no date of any kind. The same post also announced that Google had started its most ambitious pre-training run yet, for Gemini 4. Since then, what Google shipped on 2026-09-02 was Gemini 3.8 Flash and 3.8 Flash Cyber; that announcement does not update the status of 3.5 Pro.
When will Gemini 3.5 Pro release?
The verified Gemini timeline leading up to 3.5 Pro.
Launched with a 1M-token context window (LMArena Elo 1501).
Released for complex, long-horizon agentic and multimodal coding.
Shipped at Google I/O to billions of users via the Gemini app, Search AI Mode and the API.
Google said the Pro model was in internal use and would arrive 'next month' (around June).
Outlets reported release dates of July 2026 and then 12 August 2026. Both passed with no release and no confirmed new date from Google. The only checkable official statement remains the 2026-07-21 blog's "currently testing with partners… broadly available as soon as it's ready". For contrast: what Google actually shipped on 2026-09-02 was Gemini 3.8 Flash and 3.8 Flash Cyber. Checked again on 2026-10-07: the newest entry in the official Gemini API release notes (2026-10-06, Nano Banana 2.1 generally available) still does not mention 3.5 Pro.
✅ Confirmed Gemini facts you can rely on
Official benchmarks for the Gemini models that have already shipped.
Gemini 3 Pro
Released Nov 18, 2025 · 1M context
- LMArena Elo · 1501
- GPQA Diamond · 91.9%
- Humanity's Last Exam · 37.5%
- MathArena Apex · 23.4%
- Context window · 1M
Gemini 3.5 Flash
GA May 19, 2026
- Terminal-Bench 2.1 · 76.2%
- MCP Atlas · 83.6%
- CharXiv Reasoning · 84.2%
- GDPval-AA · 1656 Elo
🔮 Rumored vs confirmed
What people expect from Gemini 3.5 Pro — and what Google has actually said.
2M-token context (rumored)
Widely expected to double Gemini 3 Pro's window, but Google has not confirmed Pro specs.
Deep Think reasoning (expected)
Anticipated for the Pro tier, but not officially detailed for 3.5 Pro.
Release date: two rumoured dates have both passed, and Google has never given one
The circulated dates of July 2026 and 12 August 2026 have both come and gone without a release. Both came from financial and insider channels, and Google confirmed neither. The only checkable official phrasing is "broadly available as soon as it's ready", which contains no date. Do not plan around any circulating date; this page will not chase further rumoured dates either, only official statements.
There is no Gemini 3.2
Google skipped 3.2 and moved straight to the 3.5 family — ignore any '3.2' claims.
Pricing, context window and benchmarks for Gemini 3.5 Pro are unofficial until Google publishes them. We'll update this page on launch.
While 3.5 Pro is pending: what Google's docs say today
The cards below only summarise Google's own documentation; they do not mean QCode offers these models.
New projects: 3.5 Flash-Lite or 3.8 Flash
The 2026-09-18 Gemini API release notes, which limited access to the 2.5 models, tell new projects to use the latest models: 3.5 Flash-Lite or 3.8 Flash.
Pro tier: Gemini 3.1 Pro
As of 2026-10-07 the newest Pro model on the official Gemini API models page is Gemini 3.1 Pro (gemini-3.1-pro-preview).
This page follows official sources only
Once 3.5 Pro appears on the Gemini API models page or in the release notes, this page will be updated from there. QCode does not currently offer the Gemini series; please use Claude, GPT or Chinese models (e.g. claude-sonnet-5-5, gpt-6.1-sol, glm-5.3).
Frequently asked questions
When will Gemini 3.5 Pro be released?
As of 2026-10-07 no public date turns up on the official Gemini API models page or release notes: the models page has no 3.5 Pro (its newest Pro model is Gemini 3.1 Pro), and the newest release-notes entry, dated 2026-10-06, does not mention it. The most recent checkable official wording is from the 2026-07-21 blog: "currently testing with partners… broadly available as soon as it's ready". Outlets previously reported July 2026 and then 12 August 2026; both dates passed without a release and Google confirmed neither. What Google actually shipped on 2026-09-02 was Gemini 3.8 Flash and 3.8 Flash Cyber, and that announcement does not update the status of 3.5 Pro.
Can I use Gemini 3.5 Pro on QCode now?
No. As of 2026-10-07, 3.5 Pro does not appear on the official Gemini API models page. QCode does not currently offer the Gemini series; please use Claude, GPT or Chinese models (e.g. claude-sonnet-5-5, gpt-6.1-sol, glm-5.3).
What is the difference between Gemini 3.5 Flash and 3.5 Pro?
Flash (GA May 19) is the fast, cost-efficient tier and is already available. Pro is the high-capability flagship still in preview. Flash benchmarks are public; Pro specs are not.
Is there a Gemini 3.2?
No. Google moved from the 3.1 series straight to 3.5; there is no official Gemini 3.2 release.
One key for the models QCode actually serves
QCode serves models such as claude-sonnet-5-5, claude-opus-5-5, gpt-6.1-sol, deepseek-v4-pro, glm-5.3 and kimi-k3, billed per token; each model's price is on /models.