Short answer
Gemini 4 Flash is not out, and Google hasn't announced it. As of November 8, 2026, the only Gemini 4 model in Google's Gemini API model list is Gemini 4 Argon. Google has also announced Gemini 4 Carbon, which is open only to a group of cyber defenders. Everything about Gemini 4 Flash so far comes from leaks.
- Leak, not officially confirmed: on October 27, X user Pankaj Kumar posted that Gemini 4 Flash "is in internal testing under the codename Hydrogen" (post). In early November, other leak-tracking accounts said several Hydrogen checkpoints were being tested on Google AI Studio and LMArena (OfficeChai summary, November 5).
- A model ID string, also unconfirmed: in August, Nokiapoweruser reported that the string
gemini-4-flash-previewappeared in tokenizer code in Google's public SDK repository (report). Google hasn't published that model ID. - What you can use now: Gemini 3.8 Flash (
gemini-3.8-flash), Google's current Flash model, and Gemini 4 Argon (gemini-4-argon). Both work on APITokenRouter with the same key. The Gemini 4 Flash upcoming page tracks its status.
New accounts start with a $0 balance — top up before your first request. Top up from $5. If a request returns 402 (400 on the OpenAI format) with "wallet balance is insufficient", your balance is empty — your base URL and key are fine. Top up and retry.
Before Gemini 4 Flash ships, use Gemini 3.8 Flash or Gemini 4 Argon with the same key
What the leaks say
These are leaks, not officially confirmed.
- Pankaj Kumar's post (October 27): "Gemini 4 Flash leaks: codenamed Hydrogen. Gemini 4 Flash is in internal testing under the codename Hydrogen." The post also mentions a model ID,
gemini-3.8-flash-router, that "can be related to Gemini 4 Flash". It doesn't say where the information comes from. - Early November: according to OfficeChai's summary, accounts that track AI leaks said five Hydrogen checkpoints had surfaced, with reported speeds of 300 to 500 tokens per second, and that the model was being A/B tested on Google AI Studio and LMArena. These are posts on X, not published tests.
- August: Nokiapoweruser reported developers spotting
gemini-4-flash-previewin tokenizer code in Google's public SDK repository. A string in code doesn't mean that will be the model ID at launch.
We found no Gemini 4 Flash in Google's Gemini API docs, model list, changelog or blog.
What Google has confirmed about Gemini 4
| Gemini 4 Argon | Gemini 4 Carbon | Gemini 4 Flash | |
|---|---|---|---|
| Status | Announced September 30, 2026; open to developers | Announced October 28, 2026; open to a group of cyber defenders | Leaks only, not officially confirmed |
| Model ID | gemini-4-argon |
Not open to developers yet | Not announced (a leaked string: gemini-4-flash-preview) |
| Official price (per 1M tokens) | $2 input / $10 output through November 30, then $4 / $20 | $2 / $10 introductory, then $4 / $20 | Not announced |
Argon and Carbon figures come from the Gemini API model list and Google's Gemini 4 Carbon announcement. More detail: Gemini 4 Argon API and Gemini 4 Carbon: release and access.
What isn't known yet
Whether Gemini 4 Flash will ship, when, its final name and model ID, its price, how much faster or better it is than Gemini 3.8 Flash, and whether it will come to the Gemini API. We won't guess.
Use these Gemini models today
- Gemini 3.8 Flash (
gemini-3.8-flash): Google's current Flash model, generally available since September 2, 2026. Fast and low-cost for everyday tasks. Price, limits and examples: Gemini 3.8 Flash API. - Gemini 4 Argon (
gemini-4-argon): the Gemini 4 model open to developers now, for harder, multi-step work. See Gemini 4 Argon API.
Both use the same APITokenRouter key and base URL, in the OpenAI-compatible format (https://apitokenrouter.com/v1) or Google's native Gemini format.
If Gemini 4 Flash ships: if it becomes available on APITokenRouter, change only model in your request to the ID Google publishes. The base URL and key stay the same.
We'll update this page as soon as Gemini 4 Flash ships
When Google announces Gemini 4 Flash, we'll update this article with the official model ID, the official price, and whether it's available on APITokenRouter. You can also vote for it on the Gemini 4 Flash upcoming page.
Paying for usage
No subscription: you top up a balance from $5, and each request is charged by the tokens it uses, by card or with USDT, USDC or ETH. The same key also works for Claude and GPT models.
FAQ
Is Gemini 4 Flash out?
No. As of November 8, 2026, Google hasn't announced Gemini 4 Flash, and it isn't in the Gemini API. There are only leaks on X and a string in SDK code.
When will Gemini 4 Flash be released?
Google hasn't given a date or confirmed the model. We'll update this article when it ships.
What is Gemini 4 Hydrogen?
"Hydrogen" is the codename leakers on X use for a Flash-tier Gemini 4 model in testing. Google hasn't confirmed the codename or the model.
Is gemini-4-flash-preview the model ID?
Not confirmed. The string was reported in SDK tokenizer code in August; Google hasn't published any Gemini 4 Flash model ID.
Can I use Gemini 4 Flash now?
No. Use Gemini 3.8 Flash (gemini-3.8-flash) for fast, low-cost tasks or Gemini 4 Argon (gemini-4-argon) for harder ones.