Google · Available now
Gemini 3.8 Flash (2026): price, specs, release date, and the names people mix it up with
Gemini 3.8 Flash is Google’s Flash model for long-horizon software engineering and autonomous agents, released on September 2, 2026 with a 1M-token context window and priced at $0.75 per million input and $3.75 per million output tokens through December 31, 2026, and Google now calls it a previous-generation model next to Gemini 4 Hydrogen.
Use it now
Chat
≈ 1 credit per reply · See all models
New accounts get 5 free credits once, valid for 7 days. They work on the lower-priced models here; the priciest ones need credits you bought.
Start chatting — 5 free credits- Maker
- Type
- Text
- Released
- 2026-09-02
- OpenRouter id
- google/gemini-3.8-flash
Per million tokens
Price
| - | Input | Output | Source |
|---|---|---|---|
| Google — Standard, through December 31, 2026 | $0.75 | $3.75 | aistudio.google.com |
| Google — Standard, from January 1, 2027 | $1.5 | $7.5 | aistudio.google.com |
| Via OpenRouter (google/gemini-3.8-flash) | $0.82 | $4.125 | openrouter.ai |
Checked
Facts, with sources
- Google describes Gemini 3.8 Flash as its most intelligent Flash model, built for long-horizon software engineering, autonomous agents and complex enterprise workflows. Google’s model index now labels it "Stable" and a "previous-generation Flash model".Source: aistudio.google.com · checked 2026-11-18
- Google’s model page lists the code gemini-3.8-flash, text, image, video, audio and PDF input with text output, a 1,048,576-token input limit and a 65,536-token output limit.Source: aistudio.google.com · checked 2026-11-18
- It supports low, medium (the default) and high thinking levels; Google says the minimal level is not supported and returns an error.Source: aistudio.google.com · checked 2026-11-18
- Google’s pricing page lists Standard at $0.75 per million input tokens and $3.75 per million output tokens through December 31, 2026, and $1.50 / $7.50 from January 1, 2027. Batch and Flex are half of that.Source: aistudio.google.com · checked 2026-11-18
- Google’s model page gives only "Latest update: September 2026" and no launch day; OpenRouter lists the model from September 2, 2026. Google says that from December 1, 2026 requests to the deprecated gemini-2.5-pro are routed to gemini-3.8-flash.Source: aistudio.google.com · checked 2026-11-18
- Listed on OpenRouter as google/gemini-3.8-flash at $0.825 / $4.125 per million input / output tokens, context 1,048,576 tokens (checked November 18, 2026).Source: openrouter.ai · checked 2026-11-18 · second-hand
- Google’s Gemini API changelog lists "Gemini 3.8 Flash generally available (GA)" on September 2, 2026, and the model card carries the same date. The deprecations page lists gemini-3.8-flash from September 2, 2026 with no shutdown date announced.Source: aistudio.google.com · checked 2026-11-19
- Google says gemini-3.7-flash is deprecated: requests to it are automatically routed to gemini-3.8-flash (changelog entry of October 8, 2026), and from December 1, 2026 requests to the deprecated gemini-2.5-pro are billed at Gemini 3.8 Flash rates.Source: aistudio.google.com · checked 2026-11-19
- Google’s model card says Gemini 3.8 Flash is the next iteration of the Gemini 3 family and is based on Gemini 3.7 Flash. It takes text, images, audio and video with a context window of up to 1M tokens and outputs text with 64K tokens; the knowledge cutoff is March 2026, though Google says some domains may be limited to January 2025.Source: deepmind.google · checked 2026-11-19
- Google’s model page lists these as supported: code execution, computer use, context caching, file search, function calling, grounding with Google Maps and Google Search, structured outputs and URL context; Batch API, Flex and Priority inference are available. Not supported: audio generation, image generation and the Live API.Source: aistudio.google.com · checked 2026-11-19
- Google’s model card says Gemini 3.8 Flash is designed for cost-effective scaling of general-purpose, production-ready agents, with software engineering, agent tasks and complex knowledge workflows as example uses.Source: deepmind.google · checked 2026-11-19
- Google’s guide says the model can use more tokens on longer, complex tasks by design, because it takes smaller reasoning steps, calls tools iteratively and verifies its work; for everyday tasks it suggests lowering the reasoning effort, and says Gemini 3.6 Flash remains supported.Source: aistudio.google.com · checked 2026-11-19
- Google lists the usual limits in the model card: hallucinations, occasional slowness or timeouts, and higher token use at higher effort levels. It also says safety performance across non-English languages regressed slightly relative to Gemini 3.7 Flash.Source: deepmind.google · checked 2026-11-19
- Google’s model card lists these channels for Gemini 3.8 Flash: the Gemini app, Gemini Enterprise Agent Platform, Google AI Studio, the Gemini API, Google AI Mode and Google Antigravity.Source: deepmind.google · checked 2026-11-19
- Google’s Gemini 4 Hydrogen guide calls Hydrogen the recommended upgrade from Gemini 3.8 Flash and lists the migration changes: a new model ID, one shared 1,048,576-token window, a high default thinking level and no explicit caching.Source: aistudio.google.com · checked 2026-11-19
Price
Gemini 3.8 Flash price: Google’s rates before and after January 1, 2027
- Google’s pricing page marks the Standard rate of Gemini 3.8 Flash as "Free of charge" on the free tier of the Gemini API, with the note that on the free tier Google uses the content to improve its products (on the paid tier it does not). Batch is not offered on the free tier.Source: aistudio.google.com · checked 2026-11-19
- OpenRouter lists google/gemini-3.8-flash at $0.825 / $4.125 per million input / output tokens and a separate google/gemini-3.8-flash:batch entry (batch endpoint only) at $0.375 / $1.875; both show a 1,048,576-token context and a 65,536-token maximum completion (checked November 19, 2026).Source: openrouter.ai · checked 2026-11-19 · second-hand
Price table
Gemini 3.8 Flash price per million tokens, by tier and date
Gemini 3.8 Flash has two Google price lists, because Google marked the current rates as introductory. The Standard rate is $0.75 input / $3.75 output through December 31, 2026 and doubles on January 1, 2027, when it matches the list price of Gemini 4 Hydrogen. OpenRouter’s row is a different price, set by OpenRouter, not by Google; do not mix the two. Output prices include thinking tokens. For the whole Google family see Google Gemini.
| Tier | Through December 31, 2026 | From January 1, 2027 |
|---|---|---|
| Standard, input / output | $0.75 / $3.75 | $1.50 / $7.50 |
| Standard, cached input | $0.075 | $0.15 |
| Batch and Flex, input / output | $0.375 / $1.875 | $0.75 / $3.75 |
| Priority, input / output | $1.35 / $6.75 | $2.70 / $13.50 |
| OpenRouter list (what this site’s chat runs on), input / output | $0.825 / $4.125 (batch entry $0.375 / $1.875) | No date change on the list we read |
How good
Gemini 3.8 Flash benchmarks: Google’s own table
- Google’s own figures, not tested by us: the model card (published September 2, 2026) compares Gemini 3.8 Flash with Gemini 3.7 Flash, Claude Opus 5, Claude Sonnet 5, GPT-5.6 Sol and GPT-5.6 Terra. Gemini 3.8 Flash scores 73.7% on DeepSWE v1.1 (Claude Opus 5: 74.0%) and 19.1% on Terminal-bench 4.0 (Claude Opus 5: 51.8%, GPT-5.6 Sol: 37.3%).Source: deepmind.google · checked 2026-11-19
- Google’s own figures, not tested by us: the Gemini 4 Hydrogen announcement says Hydrogen scores 57% on Terminal-Bench 4.1 against 18.2% for Gemini 3.8 Flash, and finishes tasks about 2.5 times faster at its high default setting.Source: blog.google · checked 2026-11-19
Side by side
Gemini 3.8 Flash against five other models, as printed on Google’s model card
This is a table Google printed, with Google’s choice of rivals, so read it as a maker’s claim. It shows Gemini 3.8 Flash ahead of Gemini 3.7 Flash on every benchmark row and level with Claude Opus 5 on the two coding rows, at a fraction of the price; it also shows Claude Opus 5 well ahead on Terminal-bench 4.0, OSWorld-2.0 and GDPVal-AA v2, and GPT-5.6 Sol well ahead on Terminal-bench 4.0. Google’s newer comparison with Hydrogen is on Gemini 4 Hydrogen vs Gemini 3.8 Flash.
| Row | Gemini 3.8 Flash | Gemini 3.7 Flash | Claude Opus 5 | Claude Sonnet 5 | GPT-5.6 Sol | GPT-5.6 Terra |
|---|---|---|---|---|---|---|
| Input price | $0.75 ($1.50 regular) | $0.75 ($1.50 regular) | $5.00 | $2.00 | $4.00 | $2.00 |
| Output price | $3.75 ($7.50 regular) | $3.75 ($7.50 regular) | $25.00 | $10.00 | $20.00 | $12.00 |
| DeepSWE v1.1 (long-horizon software engineering) | 73.7% | 65.3% | 74.0% | 53.8% | 72.7% | 69.6% |
| Terminal-bench 2.1 (agentic terminal coding) | 89.4% | 85.8% | 89.1% | 80.4% | 88.8% | 87.4% |
| Terminal-bench 4.0 (general agent capabilities) | 19.1% | 11.2% | 51.8% | 12.4% | 37.3% | 23.6% |
| GDPVal-AA v2 (knowledge work, Elo) | 1545 | 1482 | 1824 | 1584 | 1710 | 1528 |
| OSWorld-2.0 (agentic computer use, partial score) | 59.0% | 50.6% | 75.4% | 42.6% | 62.6% | 50.2% |
| HLE-Verified (multidisciplinary expert reasoning) | 54.9% | 53.6% | 54.4% | 31.0% | 54.5% | 51.1% |
Names
Gemini 3.8 Flash, Gemini 3 Flash Live, Gemini 4 Helix: which names Google lists
- Google’s model index (read November 19, 2026) lists four Gemini 4 models: Gemini 4 Carbon, Gemini 4 Neon, Gemini 4 Hydrogen and Gemini 4 Argon. It has no Gemini 4 Helix, and a search of Google’s docs index (llms.txt) for "helix" found nothing.Source: aistudio.google.com · checked 2026-11-19
- The same index lists Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking for the Live API, a legacy Gemini 3.1 Flash Live preview ("we recommend updating to Gemini 3.8 Live") and a Gemini 3 Flash preview. No model there is called Gemini 3 Flash Live.Source: aistudio.google.com · checked 2026-11-19
- Google’s model page for Gemini 3.8 Live (gemini-3.8-live) lists text, image and audio input, text and audio output, an input limit of 131,072 tokens, an output limit of 65,536 tokens and Live API support (latest update September 2026).Source: aistudio.google.com · checked 2026-11-19
- OpenRouter’s model list has no Gemini Live model (no id with "live" in it among its Google entries; checked November 19, 2026).Source: openrouter.ai · checked 2026-11-19 · second-hand
Names checked
Gemini names people search for, and what Google’s model index says
Searches for "gemini 3 flash live" and "gemini 4 helix" land on this page, but only one of the names is a Google model that sits near Gemini 3.8 Flash. The table puts every name next to what Google’s own index says. Where we found no Google page, the row says so. The real Gemini 4 models have their own pages: Gemini 4 Carbon, Gemini 4 Hydrogen and Gemini 4 Argon.
| Name | On Google’s index? | Model code | What Google says |
|---|---|---|---|
| Gemini 3.8 Flash | Yes, Stable | gemini-3.8-flash | Most intelligent Flash model, for software engineering, agents and enterprise workflows; Live API not supported |
| Gemini 3.8 Live | Yes, New | gemini-3.8-live | Default low-latency audio-to-audio model for voice agents on the Live API; input limit 131,072 tokens, output limit 65,536 tokens |
| Gemini 3.8 Live Extended Thinking | Yes, New | gemini-3.8-live-extended-thinking | High-reasoning Live API model for voice, recommended when more background reasoning is needed |
| Gemini 3.1 Flash Live | Yes, Preview | gemini-3.1-flash-live-preview | Legacy Live API preview model; Google recommends updating to Gemini 3.8 Live |
| Gemini 3 Flash | Yes, Preview | gemini-3-flash-preview | Frontier-class performance rivaling larger models at a fraction of the cost |
| Gemini 4 Carbon | Yes | gemini-4-carbon | Google’s strongest reasoning model, for complex software engineering, multimodal knowledge work, scientific reasoning and cybersecurity defense |
| Gemini 3 Flash Live | No page found | None | Not in the index; the closest names are Gemini 3.8 Live, Gemini 3.1 Flash Live and Gemini 3 Flash |
| Gemini 4 Helix | No page found | None | Not in the index, which lists four Gemini 4 models: Carbon, Neon, Hydrogen and Argon |
Source: aistudio.google.comSource: aistudio.google.comSource: aistudio.google.comSource: aistudio.google.com
Nearest available
Models you can use now instead
Gemini 4 Hydrogen
Google$1.5 / $7.5 (per million tokens)
- Google announced Gemini 4 Hydrogen on November 12, 2026 (the post was last updated November 16, 2026), and its model card is dated the same day.Source: blog.google · checked 2026-11-17
- The Gemini API changelog lists gemini-4-hydrogen as generally available (GA) on November 12, 2026, with a new xhigh thinking level and as the default model for the Antigravity agent in Gemini Managed Agents; it points to migration guidance from Gemini 3.8 Flash.Source: aistudio.google.com · checked 2026-11-17
- Google describes Gemini 4 Hydrogen as a fast, efficient model for everyday coding and agentic work, built for "rapid execution across diverse agentic workflows". Gemini 4 Carbon and Gemini 4 Argon remain its frontier models for long-horizon reasoning.Source: blog.google · checked 2026-11-17
FAQ
Questions
What is Gemini 3.8 Flash?
Google describes it as its most intelligent Flash model, built for long-horizon software engineering, autonomous agents and complex enterprise workflows. Its model card says it is based on Gemini 3.7 Flash. It takes text, image, video, audio and PDF input and outputs text. Google’s model index now labels it a previous-generation Flash model, with Gemini 4 Hydrogen as the recommended upgrade.
When was Gemini 3.8 Flash released?
Google’s changelog lists it as generally available on September 2, 2026, and its model card is dated the same day; OpenRouter lists it from the same date. Since then Google has routed requests for the deprecated gemini-3.7-flash to it (October 8, 2026), and from December 1, 2026 it does the same for the deprecated gemini-2.5-pro.
How much does Gemini 3.8 Flash cost?
Google’s pricing page lists Standard at $0.75 per million input and $3.75 per million output tokens through December 31, 2026, then $1.50 and $7.50 from January 1, 2027; Batch and Flex are half of that and Priority is $1.35 / $6.75, rising to $2.70 / $13.50. The Standard rate has a Gemini API free tier, where Google uses your content to improve its products. OpenRouter’s list is a separate price, $0.825 / $4.125 (checked November 19, 2026). On this site a typical reply to this model costs about 2 credits (a credit is about $0.01 in the $5 pack), billed by its real token cost, at least 1 credit.
What is the Gemini 3.8 Flash context window, and what can it do?
Google’s model page lists a 1,048,576-token input limit and a 65,536-token output limit, with text, image, video, audio and PDF input and text output. Thinking levels are low, medium (the default) and high; minimal returns an error. It supports function calling, structured outputs, code execution, computer use, file search, URL context, and Google Search and Maps grounding, but not audio or image generation or the Live API. The model card gives a March 2026 knowledge cutoff.
Is there a Gemini 3 Flash Live?
We found no Google page for that name. Google’s model index lists Gemini 3.8 Live (gemini-3.8-live), Gemini 3.8 Live Extended Thinking, a legacy Gemini 3.1 Flash Live preview and a Gemini 3 Flash preview. Gemini 3.8 Flash itself is not a voice model: Google’s page marks the Live API as not supported. OpenRouter’s list has no Gemini Live model, so this site cannot run one.
Is there a Gemini 4 Helix, and what is Gemini 4 Carbon?
We found no Google page for Gemini 4 Helix; Google’s index lists Gemini 4 Carbon, Neon, Hydrogen and Argon. Carbon is Google’s strongest reasoning model, for complex software engineering, multimodal knowledge work, scientific reasoning and cybersecurity defense. See Gemini 4 Carbon, Gemini 4 Argon and the Google Gemini overview.
Should I use Gemini 3.8 Flash or Gemini 4 Hydrogen?
Google recommends Hydrogen as the upgrade and lists the migration changes (a new model ID, one shared 1M window, a high default thinking level, no explicit caching). Google’s own figures, which we did not test, are 57% on Terminal-Bench 4.1 against 18.2% for 3.8 Flash and about 2.5 times faster tasks. Price is the other half: 3.8 Flash is cheaper through 2026, and from January 1, 2027 both list $1.50 / $7.50. See the side-by-side.
Is this the official Google website?
No. This page is unofficial and not affiliated with or endorsed by Google. Facts link to the pages we read.