Roar AI
DiscordGet started
All models
Google · Chat

Gemini 4 Neon API in Sri Lanka

Gemini 4 Neon costs LKR 999 per million input tokens and LKR 4,995 per million output tokens on Roar AI, billed in Sri Lankan rupees on one invoice. Google’s balanced Gemini 4, first in independent finance and legal tests, and slow to start.

InputLKR 999per 1M tokens
OutputLKR 4,995per 1M tokens
Context window1.05Mtokens

Gemini 4 Neon API pricing in Sri Lanka

InputLKR 999 per 1M tokens · $3
OutputLKR 4,995 per 1M tokens · $15
Cached inputLKR 49.95 per 1M tokens · $0.15
Batch inputLKR 500 per 1M tokens · $1.50
Batch outputLKR 2,498 per 1M tokens · $7.50
Priority inputLKR 1,798 per 1M tokens · $5.40
Priority outputLKR 8,991 per 1M tokens · $27

Prices are shown in LKR to make model costs easier to estimate and manage for Sri Lankan teams, with the published US dollar rate beside each.

Example: 1,000 requests with 3,000 input and 1,000 output tokens each cost LKR 7,992. Every rate is published and stays the same from one request to the next. See every model.

Rates as of 20 November 2026. Rupee prices are our published US dollar rates converted at LKR 333 to the dollar, our rate for November 2026 — the same rate your invoice uses.

Specs

Made byGoogle
ReleasedNovember 2026
TypeChat
Context window1.05M tokens
Longest answer66K tokens
ReadsText, images, PDFs
Thinks before answeringAlways on
WeightsClosed
Processed inGlobal
Model idgemini-4-neon

Is Gemini 4 Neon worth the price?

For finance, legal and long-document work, it is the one to test first. Vals AI ranked it first on both its finance and legal indexes.

For quick answers and live chat, no: it thinks before every reply and takes tens of seconds to start. Use a lower effort setting, or a faster model. For agentic coding, compare it with Claude Opus 5.5, which Google’s own table puts level with it.

Gemini 4 Neon and the models to compare it with, priced per 1M tokens
ModelInputOutputChoose it for
Gemini 4 NeonLKR 999LKR 4,995Finance, legal and long documents.
Gemini 4 ArgonLKR 666LKR 3,330Research work where a refusal beats a wrong answer.
Claude Opus 5.5LKR 1,332LKR 6,660Long coding sessions and agentic coding.
Gemini 3.8 FlashLKR 250LKR 1,249Faster answers, with video and audio input.

Prices per 1M tokens.

About Gemini 4 Neon

Neon is the Gemini 4 model Google describes as balancing depth and efficiency, and the one Google suggests in place of Gemini 3.1 Pro. It was released on 18 November 2026. Independent tests rate it highly: Vals AI placed it second of all the models it has tested on the Vals Index and first on its finance and legal indexes, and Artificial Analysis scored it 61 on its Intelligence Index, ahead of Claude Opus 5.5.

It reads text, images and PDFs in a window of about a million tokens, and it always thinks before answering, from low to extra-high effort. Google counts that thinking as output, so even a short reply is billed for the thinking behind it.

We measured what its thinking costs at each effort setting, through our API on launch week: GPT-6.2 Astra, GPT-6.2 Sol and Gemini 4 Neon: pricing and API tests.

What is Gemini 4 Neon good for?

  • Finance and legal analysis across long documents.
  • Questions that need facts found and connected across a very long input.
  • Research and science work where the answer can wait.

Gemini 4 Neon limitations

  • Slow to start. At its default effort, Artificial Analysis measured about 40 seconds before the first word of the answer, and about 14 seconds at the lowest. Not a fit for live chat.
  • Wordy at its highest effort: on Artificial Analysis’s tests it wrote over four times the median number of tokens, which adds to what a task costs.
  • Google’s own table puts it level with Claude Opus 5.5 on agentic coding rather than ahead.
  • Video and audio input, and Google’s very long answers, are not available through our API yet; answers are capped at about 65K tokens.

Using Gemini 4 Neon from Sri Lanka

Roar AI provides access to Gemini 4 Neon through the Roar AI gateway, with usage billed in LKR and no foreign card needed. Developers can use Gemini 4 Neon alongside other supported models without maintaining a separate integration for every provider.

Google publishes no Sinhala or Tamil test results for Neon. If your users write in either language, try it on real messages before you build on it.

Use the OpenAI SDK you already have: point it at Roar AI and set the model name. Read the docs.

from openai import OpenAI

client = OpenAI(base_url="https://api.roar-ai.com/v1", api_key="roar_live_…")

reply = client.chat.completions.create(
    model="gemini-4-neon",
    messages=[{"role": "user", "content": "Summarise this contract in plain English."}],
)

Where is Gemini 4 Neon processed?

Gemini 4 Neon is a frontier model, and requests are processed on international infrastructure, outside Sri Lanka.

Roar AI’s locally hosted model is separate and is labelled as such: if a workload has to stay in the country, Qwen 3.8 27B LK runs on our own server in Colombo.

Gemini 4 Neon FAQ

How much does Gemini 4 Neon cost in Sri Lanka?

Input is LKR 999 per 1M tokens; output is LKR 4,995 per 1M tokens. For example, 1,000 requests with 3,000 input and 1,000 output tokens each cost LKR 7,992. Rupee prices are our published US dollar rates converted at LKR 333 to the dollar, our rate for November 2026 — the same rate your invoice uses.

Can I pay for Gemini 4 Neon in Sri Lankan rupees?

Yes. Gemini 4 Neon usage is billed in LKR on one monthly invoice with every other model you use, so there is no foreign card or separate Google account to set up.

Is Gemini 4 Neon open source?

No. Google does not publish the weights, so it is only available through an API like this one.

What is the context window of Gemini 4 Neon?

Gemini 4 Neon can read up to 1.05M tokens in one request, and write an answer of up to 66K tokens.

Can Gemini 4 Neon read images and PDFs?

Yes. Gemini 4 Neon reads images and PDFs directly.

Start using Gemini 4 Neon

Create a free account, get an API key and call Gemini 4 Neon with the OpenAI SDK, billed in Sri Lankan rupees.