Gemini Model Release Timeline: Every Gemini Version from 1.0 to Gemini 4 Argon (2023–2026)
Live timeline · last updated · 1 entry in the update log
One dated list of Google’s Gemini language models, from the 1.0 launch to Gemini 4 Argon. Launch announcements come from Google’s blog; API preview, GA and deprecation dates come from the Gemini API changelog.
As of Oct 21, 2026 Google’s newest Gemini models are Gemini 4 Argon, generally available since Oct 13, 2026, and Gemini 3.8 Flash, generally available since Sep 2, 2026; the series began with Gemini 1.0 on Dec 6, 2023.
Read on 2026-10-21: the Gemini API changelog on ai.google.dev, Google’s Gemini 1.0 and 1.5 announcements, and both Gemini 4 Argon posts on blog.google. Dates are those printed on the pages. We list text and multimodal language models, not image, speech, embedding or Veo releases, and we ran no benchmarks.
Timeline
-
Google introduces Gemini 1.0 in three sizes, Ultra, Pro and Nano, as its first natively multimodal model family.
-
Google announces Gemini 1.5, starting with 1.5 Pro and an experimental 1 million-token context window for a limited group of developers and enterprise customers.
-
Gemini 1.5 Pro is released in preview in the Gemini API as gemini-1.5-pro-latest.
-
Gemini 1.5 Flash is released in preview as gemini-1.5-flash-latest.
-
gemini-1.5-pro-002 and gemini-1.5-flash-002 are released as new stable versions for general availability.
-
Gemini 2.0 Flash Experimental is released for public preview, which the changelog says is twice as fast as Gemini 1.5 Pro and adds the bidirectional Live API.
-
Gemini 2.0 Flash Thinking Mode is released for public preview, showing the model’s thought process as it answers.
-
Gemini 2.0 Flash becomes generally available as gemini-2.0-flash-001, and experimental previews of Gemini 2.0 Pro and 2.0 Flash-Lite are released.
-
Gemini 2.0 Flash-Lite becomes generally available.
-
Gemini 2.5 Pro Experimental is released, a public experimental model with thinking mode on by default.
-
Gemini 2.5 Flash is released in preview, optimized for price-performance and adaptive thinking.
-
Gemini 2.5 Pro and Gemini 2.5 Flash reach stable versions as gemini-2.5-pro and gemini-2.5-flash.
-
Gemini 2.5 Flash-Lite is released as a fast, low-cost Gemini 2.5 model.
-
Gemini 3 Pro Preview launches as the first Gemini 3 series model.
-
Gemini 3 Flash Preview launches, with upgraded visual and spatial reasoning and agentic coding.
-
Gemini 3.1 Pro Preview is released, along with a separate custom-tools endpoint.
-
Gemini 3.1 Flash-Lite Preview launches as the first Flash-Lite model in the Gemini 3 series.
-
Gemini 3.1 Flash-Lite becomes generally available.
-
Gemini 3.5 Flash becomes generally available and becomes the model behind gemini-flash-latest.
-
Gemini Omni Flash enters public preview as a multimodal model for fast video generation and conversational video editing, producing 3–10 second clips at 720p.
-
Gemini 3.6 Flash and Gemini 3.5 Flash-Lite become generally available; the changelog says 3.6 Flash is cheaper than 3.5 Flash with better token efficiency.
-
Gemini 3.7 Flash becomes generally available, described as Google’s most intelligent workhorse model yet for coding and agents, at an introductory price.
-
Gemini 3.8 Flash becomes generally available as Google’s most intelligent Flash model, aimed at long-horizon software engineering and autonomous agents.
-
Google announces Gemini 4 Argon and first makes it available to trusted cyber defenders through its Fairwind Program.
-
Gemini 3.7 Flash is deprecated; API requests to gemini-3.7-flash are automatically routed to gemini-3.8-flash.
-
Gemini 4 Argon becomes generally available, described as Google’s most capable model for complex coding, agentic, math and science tasks.
How to read this list
Google usually publishes a model in stages: an experimental or preview model in the Gemini API and AI Studio, then a stable, generally available (GA) version, and later a deprecation that routes the old model name to a new one. We record the stage named in each entry, so the same model can appear twice, as Gemini 2.5 Pro does in March and June 2025. Two early entries, Gemini 1.0 and the Gemini 1.5 announcement, use Google’s blog posts because they came before the API launches.
The changelog gives a calendar day for each entry but no time of day or time zone. We copy the day as printed. If you are in Asia and a model seemed to appear a day later than listed here, that is likely the time zone difference rather than a different release; we have not tried to convert days we cannot time. The same applies to the two blog posts about Gemini 4 Argon, which are dated by day only.
We left out image models such as the Nano Banana line, speech models such as Gemini 3.5 Transcribe, embedding models and Veo. Each has its own changelog entries and would make this list hard to scan.
The naming, in one paragraph
Each Gemini generation has come in sizes. Gemini 1.0 had Ultra, Pro and Nano. From 1.5 onward the main sizes are Pro (largest), Flash (fast workhorse) and Flash-Lite (cheapest). In 2026 Google released Flash versions faster than Pro versions: 3.5, 3.6, 3.7 and 3.8 Flash all reached GA between May and September, while the last Pro entry in the changelog is 3.1 Pro Preview from February. Gemini 4 broke the pattern with a name, Argon, rather than a size.
2026 in detail: the Flash run and Gemini 4 Argon
From May to September 2026 a new Flash model reached general availability roughly every four to seven weeks: 3.5 Flash on May 19, 3.6 Flash on July 21, 3.7 Flash on August 13 and 3.8 Flash on September 2. The turnover was fast enough that 3.7 Flash was deprecated on October 8, less than two months after its GA, with requests routed to 3.8 Flash. If you pin model names in code, that routing behavior is the main practical takeaway.
Gemini 4 Argon came in two steps. On September 30 Google announced it and made it available first to trusted cyber defenders through the Fairwind Program. On October 13 it became generally available. Our separate Google Argon timeline covers that rollout in more detail, and what you can do with Google Argon today covers access.
Model names that moved under developers
Google has repeatedly pointed old or alias model names at newer models, which matters more for production code than any launch date. The changelog records several cases, and each one is a reason to pin a dated model string and to read the changelog before upgrading.
- September 24, 2024: gemini-1.5-pro-latest and gemini-1.5-flash-latest were updated to point to the new -002 stable versions.
- June 17, 2025: the changelog said gemini-2.5-pro-preview-05-06 would be redirected to gemini-2.5-pro on June 26, 2025.
- May 7, 2026: with Gemini 3.1 Flash-Lite at GA, the preview model was set to be shut down on May 25, 2026.
- May 19, 2026: Gemini 3.5 Flash became the model behind the gemini-flash-latest alias.
- October 8, 2026: requests to gemini-3.7-flash started being routed automatically to gemini-3.8-flash, and 3.7 Flash was removed from the Google AI Studio model picker.
Which Gemini model is the newest, by tier
- Most capable: Gemini 4 Argon, generally available October 13, 2026.
- Flash: Gemini 3.8 Flash, generally available September 2, 2026.
- Flash-Lite: Gemini 3.5 Flash-Lite, generally available July 21, 2026.
- Video: Gemini Omni Flash, in public preview since June 30, 2026.
What is not verified here
- Release dates for other Gemini 4 sizes. Google has not announced them in the sources we read.
- The Gemini app rollout date for each model by country and plan. App rollouts are staged and not all are dated.
- Benchmarks and parameter counts. Google has published benchmarks for several models; we did not reproduce them and cite none as facts.
Related timelines
Compare the Claude model timeline, the OpenAI GPT model timeline and the Grok model timeline. The Front blog index lists every post.
The same format also covers Moonshot Kimi, Mistral, DeepSeek and Alibaba Qwen.
Keep reading on Front
Every post, newest first, is on the Front blog index.
- Claude model release timeline: every Claude version with its release date.
- OpenAI GPT model release timeline: every GPT and o-series model from ChatGPT to GPT-6.
- Grok model release timeline: every Grok version, with Elon Musk’s posts quoted and linked.
- Google Argon timeline: the dated record of Google’s Gemini 4 Argon launch.
- AI model releases in 2024: the 2024 nodes from the vendor timelines, in date order.
- AI model releases in 2025: the 2025 nodes from the vendor timelines, in date order.
- AI model releases in 2026: the 2026 nodes from the vendor timelines, in date order.
- Qwen model timeline: Qwen from Qwen-7B to Qwen3-Coder, with licenses and context.
FAQ
What is the latest Gemini model?
As of October 21, 2026: Gemini 4 Argon, generally available since October 13, 2026. The newest Flash model is Gemini 3.8 Flash, generally available since September 2, 2026.
When was Gemini 4 released?
Gemini 4 Argon was announced on September 30, 2026 for trusted cyber defenders and became generally available on October 13, 2026.
When was Gemini 3 released?
Gemini 3 Pro Preview launched on November 18, 2025 and Gemini 3 Flash Preview on December 17, 2025.
When was the first Gemini released?
Google introduced Gemini 1.0 on December 6, 2023.
Update log
- — First published. Checked Google’s blog and the Gemini API changelog through Oct 21, 2026 (UTC+8).