TokenMix Research Lab · 2026-05-22

Gemini 3.5 Pro Status: Coming Soon, 3.1 Pro Preview Live Now

Last verified: 2026-05-22 Author: TokenMix Research Lab

Gemini 3.5 Pro is not released as of 2026-05-22. Google DeepMind's official Gemini models page labels it "3.5 Pro coming soon" with no published date. The current Gemini Pro tier in production is Gemini 3.1 Pro Preview (API ID gemini-3.1-pro-preview), and the only Gemini 3.5-line model that ships today is Gemini 3.5 Flash (gemini-3.5-flash, Stable).

This page is a status reference. Every claim below carries a verified date and links to the Google source that confirms it. When Google ships Gemini 3.5 Pro, this page will be updated with the API ID, pricing, and release date on the same day.

Table of Contents

• Is Gemini 3.5 Pro Released?
• Gemini 3.x Family: What Ships Today vs Coming Soon
• Gemini 3.1 Pro Preview: The Current Pro Tier
• Gemini 3.5 Flash: The Only Stable 3.5 Model
• Pricing Across the Gemini 3.x Family
• Where to Watch for Gemini 3.5 Pro Release
• FAQ
• Sources
• TokenMix Take

Is Gemini 3.5 Pro Released?

No. Gemini 3.5 Pro is not released as of 2026-05-22. It does not appear in Google's Gemini API models documentation, it is not listed on Google's Gemini API pricing page, and no gemini-3.5-pro model ID exists in either AI Studio or Vertex AI's published catalogs. Google DeepMind's Gemini models page explicitly labels it "3.5 Pro coming soon."

If you are searching for "Gemini 3.5 Pro API," the closest currently-shipping product is Gemini 3.1 Pro Preview. If you are searching for a stable Gemini 3.5-line model, only Gemini 3.5 Flash exists today, and it is the Flash tier (faster, cheaper, less capable than the Pro tier will be).

Gemini 3.x Family: What Ships Today vs Coming Soon

The current Gemini 3.x lineup spans Pro (Preview only), Flash (one Stable, one Preview), and the still-unreleased 3.5 Pro flagship. Status reflects Google's API models docs as of 2026-05-22.

Two practical consequences: (a) within the Gemini 3.x family there is no Pro-tier model in Stable status — every 3.x Pro model is either Preview or shut down. (b) Per Google's API models docs: "Gemini 3 Pro Preview is deprecated and has been shut down March 9, 2026. Migrate to Gemini 3.1 Pro Preview to avoid service disruption." Note that Gemini 2.5 Pro from the prior generation remains available at gemini-2.5-pro per Google's API pricing page — it has not been shut down. This page focuses on the Gemini 3.x line; if you need a current production Pro model and prefer not to use Preview status, Gemini 2.5 Pro is the alternative.

Gemini 3.1 Pro Preview: The Current Pro Tier

Gemini 3.1 Pro Preview is the production Pro tier today, priced at $2 input / $12 output per 1M tokens for prompts under 200K (Standard tier, paid). API ID is gemini-3.1-pro-preview, with a variant gemini-3.1-pro-preview-customtools for custom-tool workflows. Capabilities per Google's docs: "advanced intelligence, complex problem-solving skills, and powerful agentic and vibe coding capabilities."

Pricing scales for longer prompts: input rises to $4 per MTok and output to $18 per MTok when the prompt exceeds 200K tokens. Confirmed on Google's pricing page, 2026-05-22. Batch and Flex tiers are 50% off Standard. Free tier is not available for this model.

Gemini 3.5 Flash: The Only Stable 3.5 Model

Gemini 3.5 Flash is the only Gemini 3.5-line model in Stable status. API ID is gemini-3.5-flash. Google describes it as "most intelligent model for sustained frontier performance on agentic and coding tasks." Standard pricing is $1.50 input / $9 output per 1M tokens; Batch and Flex drop to $0.75 / $4.50. Confirmed 2026-05-22.

A separate model, gemini-3-flash-preview, exists at lower Standard pricing ($0.50 input / $3 output per MTok for text), but it is Preview status — not recommended for production until Stable.

Pricing Across the Gemini 3.x Family

All Standard-tier paid pricing per 1M tokens, sourced from Google's Gemini API pricing page. All Verified 2026-05-22.

Two tier discounts available on most models: Batch = 50% off Standard (24-hour completion), Flex = 50% off Standard (latency-tolerant), Priority = 80% premium over Standard (where available). Context cache storage costs $1.00/MTok per hour on Flash and $4.50/MTok per hour on Pro.

Grounding with Google Search is included free for the first 5,000 queries per month across all Gemini 3.x models (shared quota), then $14 per 1,000 search queries.

Where to Watch for Gemini 3.5 Pro Release

Google has not announced a Gemini 3.5 Pro release date. The three primary channels to monitor are: (a) the Google DeepMind Gemini models page, which currently displays the "3.5 Pro coming soon" badge — the badge will switch to "Try in Gemini" on launch day. (b) The Google AI Gemini API models documentation, which adds new API model IDs the same day a model ships. (c) The Google AI Gemini API pricing page for confirmed input/output rates.

Historical cadence: Gemini 2.5 → Gemini 3.x took roughly 6 months. The 3.1 Pro Preview → 3.5 Pro gap has been actively narrowing as Google ships incremental 3.x updates. No public Google statement commits to a specific quarter.

FAQ

Is Gemini 3.5 Pro released?

No. As of 2026-05-22, Gemini 3.5 Pro does not exist as a callable API model. Google DeepMind labels it "coming soon" with no published date.

What is the current Gemini Pro model in production?

Within the Gemini 3.x family, Gemini 3.1 Pro Preview (gemini-3.1-pro-preview) is the active Pro model as of 2026-05-22, per Google's API models docs. It is Preview status, not Stable. From the prior generation, Gemini 2.5 Pro (gemini-2.5-pro) also remains available and is the alternative if you prefer a non-Preview production target.

What is the API model ID for Gemini 3.5 Pro?

No model ID has been assigned. Google has not published a gemini-3.5-pro model ID in AI Studio, Gemini API, or Vertex AI as of 2026-05-22.

How much will Gemini 3.5 Pro cost?

Google has not published Gemini 3.5 Pro pricing. The current Pro-tier reference point is Gemini 3.1 Pro Preview at $2/$12 per 1M tokens for prompts ≤200K, $4/$18 for prompts >200K.

Where is Gemini 3.5 Pro documented?

It is not documented in Google's developer surfaces. The "3.5 Pro coming soon" marker on DeepMind's models page is currently the only acknowledgment that the model is in the pipeline.

Can I use Gemini 3.5 Pro in Vertex AI?

No. Vertex AI's models documentation lists Gemini 3.1 Pro Preview and Gemini 3.5 Flash, but no Gemini 3.5 Pro entry exists as of 2026-05-22.

What's the difference between Gemini 3.1 Pro and Gemini 3.5 Flash?

Gemini 3.1 Pro Preview is the Pro tier — higher reasoning quality, higher cost ($2-$4 input / $12-$18 output per MTok). Gemini 3.5 Flash is the Flash tier — faster, cheaper ($1.50 input / $9 output), positioned for agentic and coding workloads. Both are documented at Google's API models page.

Was Gemini 3.5 Pro announced at Google I/O 2026?

No. Google I/O 2026 keynote shipped Gemini 3.5 Flash as Stable, not Gemini 3.5 Pro. The Pro tier remained at Gemini 3.1 Pro Preview after the keynote.

Should I wait for Gemini 3.5 Pro or use Gemini 3.1 Pro Preview now?

Use Gemini 3.1 Pro Preview if you need Pro-tier quality today. Wait for 3.5 Pro only if your workload can tolerate an undefined timeline. There is no public Google commitment to a release date.

Is Gemini 2.5 Pro still available?

Yes. Gemini 2.5 Pro remains listed on Google's API pricing page with the model ID gemini-2.5-pro. The model that was shut down on 2026-03-09 is Gemini 3 Pro Preview (gemini-3-pro-preview), per Google's docs, with migration directed to Gemini 3.1 Pro Preview.

Where will Gemini 3.5 Pro launch first — Gemini app, API, or Vertex?

Google has not stated. Historical pattern across Gemini 3.x: consumer Gemini app and AI Studio typically ship the same day, with Vertex AI following within 1-2 weeks.

How long has the "coming soon" label been on the DeepMind page?

This page does not track that history. Verify the current status directly at deepmind.google/models/gemini — the label persists until launch day.

Sources

All facts on this page sourced from Google-owned domains, verified 2026-05-22:

• Google DeepMind — Gemini models (status badges, "3.5 Pro coming soon")
• Google AI — Gemini API models documentation (API IDs, capabilities, status)
• Google AI — Gemini API pricing (Standard / Batch / Flex / Priority tier rates)
• Google Cloud — Vertex AI generative AI models (Vertex availability)

TokenMix Take

Editorial section. The TokenMix research team's interpretation, separated from the verified facts above.

The "coming soon" badge on a Pro-tier model is unusually transparent for Google — it implies an active rollout pipeline rather than a vague future intent. The asymmetry of Gemini 3.5 Flash being Stable while 3.5 Pro is unannounced is itself a data point: it suggests that distillation or specialization paths reached Flash first, and the larger 3.5 Pro is going through extended safety or capacity scaling.

For developers building on the Pro tier today, the practical recommendation is to architect against gemini-3.1-pro-preview and treat the model parameter as a config value, not a hardcoded string. When Gemini 3.5 Pro launches, a single environment variable change should be enough to migrate.

TokenMix.ai routes Gemini models through an OpenAI-compatible endpoint, so the same migration pattern works without an account in Google Cloud — the model parameter still controls which Gemini variant runs. This page will be PUT-updated within 24 hours of the official Gemini 3.5 Pro release with the new API ID, pricing, and a Verified date stamp.