Developers and AI researchers often need a powerful, multilingual model that can be accessed through a familiar OpenAI‑compatible endpoint. Gemma 3 12B is one of the 40 active models on the LLM Resayil Portal (https://llm.resayil.io). The portal delivers OpenAI‑ and Anthropic‑compatible API access, Arabic language support, streaming, function calling, vision, tool use, and a simple pay‑per‑use credit model billed in USD via Stripe or PayPal. This guide walks you through everything you need to know to start using Gemma 3 12B in your applications.

Introduction

Developers and AI researchers often need a powerful, multilingual model that can be accessed through a familiar OpenAI‑compatible endpoint. Gemma 3 12B is one of the 40 active models on the LLM Resayil Portal (https://llm.resayil.io). The portal delivers OpenAI‑ and Anthropic‑compatible API access, Arabic language support, streaming, function calling, vision, tool use, and a simple pay‑per‑use credit model billed in USD via Stripe or PayPal. This guide walks you through everything you need to know to start using Gemma 3 12B in your applications.

Comparison Table

| Feature | LLM Resayil (Gemma 3 12B) | OpenAI API (GPT‑4o) | |---|---|---| | Model catalog | 40 models, including Gemma 3 12B | GPT‑4o family | | API compatibility | OpenAI‑compatible, Anthropic‑compatible | OpenAI‑compatible | | Arabic language support | ✅ (explicit support) | ✅ (supported but not a primary focus) | | Streaming | ✅ | ✅ | | Function calling | ✅ | ✅ | | Vision | ✅ (via vision‑enabled models) | ✅ | | Tool use / thinking models | ✅ | ✅ | | Pricing model | Pay‑per‑use credits (USD) | Tiered subscription & pay‑per‑use | | Payment methods | Stripe, PayPal | Credit card, Azure billing | | Hosting location | USA | Multiple regions | | Integrations | n8n, LangChain, LiteLLM, OpenAI SDK, Anthropic SDK, Python, JavaScript, cURL | Official OpenAI SDKs, LangChain, Azure Functions |

What LLM Resayil Offers

LLM Resayil provides a single API surface that works with the same request format used by OpenAI and Anthropic. All 40 models—including Gemma 3 12B—inherit the platform’s core features:

  • Arabic language support – generate, translate, and analyse Arabic text with the same quality as English.
  • Multilingual capability – the model understands and produces many languages out of the box.
  • Streaming responses – get token‑by‑token output for real‑time UI updates.
  • Function calling – define JSON‑structured functions and let the model invoke them automatically.
  • Vision – vision‑enabled models can process images (e.g., Qwen3‑VL 235B) alongside text.
  • Thinking models & tool use – combine chain‑of‑thought prompting with external tools for complex workflows.
  • Pay‑per‑use credits – you purchase only the compute you consume, billed in USD.
  • OpenAI & Anthropic SDK compatibility – reuse existing client libraries without code changes.

These capabilities let you build sophisticated AI experiences while keeping integration effort low.

What OpenAI API (GPT‑4o) Offers

OpenAI’s GPT‑4o provides state‑of‑the‑art performance, multi‑modal capabilities, and a mature ecosystem of SDKs and tooling. It offers tiered pricing, regional availability, and extensive documentation. While GPT‑4o excels in raw performance, it does not ship with dedicated Arabic‑first‑language optimisations and requires a subscription‑based billing model.

Why LLM Resayil Wins for Gemma 3 12B Use Cases

For developers targeting Arabic‑speaking audiences or multilingual markets, Resayil’s Arabic‑first support and pay‑per‑use economics give a clear advantage. You can spin up a prototype, test traffic spikes, and only pay for the tokens you actually consume. The OpenAI‑compatible endpoint means you can migrate code between providers with minimal changes, preserving your investment in tooling like LangChain or n8n.

What You Get by Using LLM Resayil

  • Immediate access to Gemma 3 12B via /v1/chat/completions and /v1/messages.
  • Full feature set – streaming, function calling, vision (when using a vision‑enabled model), and tool integration.
  • Transparent billing – credits are deducted per token; you can top‑up at any time via /v1/pricing/topups.
  • Developer‑friendly integrations – plug‑and‑play with n8n, LangChain, LiteLLM, and the standard OpenAI/Anthropic SDKs.
  • US‑based hosting for compliance‑friendly data residency.

What Is Gemma 3 12B and Why Use It via LLM Resayil? {#what-is-gemma}

Gemma 3 12B is a chat‑optimized language model that balances size and efficiency. As one of the 40 active models on the Resayil catalog, it inherits the platform’s robust feature set:

  • OpenAI‑compatible API – call the model with the same JSON schema used for OpenAI models.
  • Arabic language support – ideal for Arabic‑centric applications.
  • Multi‑language competence – handle code‑switching and non‑Arabic content seamlessly.
  • Streaming & function calling – build responsive chat UIs and automate workflows.
  • Pay‑per‑use credits – you only purchase the compute you need, billed in USD via Stripe or PayPal.

Developers choose Gemma 3 12B on Resayil when they need a versatile, cost‑effective model that integrates instantly with existing OpenAI‑style tooling.

Key Capabilities: Multilingual, Arabic, Streaming, Function Calling, and More {#capabilities}

Multilingual & Arabic Support

The Resayil platform advertises Arabic language support and multilingual capabilities. When you send a prompt in Arabic or mix languages, Gemma 3 12B produces fluent, context‑aware output. This is especially useful for:

  • Customer‑support bots serving Arabic‑speaking regions.
  • Content generation that blends English technical terms with Arabic explanations.

Streaming Responses

Enable the stream parameter on /v1/chat/completions to receive a streaming response. Tokens are delivered as they are generated, allowing you to:

  • Show typing indicators in UI.
  • Process partial results for real‑time analytics.

Function Calling

Define functions in the request payload. Gemma 3 12B can call functions automatically, returning structured JSON. This enables:

  • Automated data look‑ups.
  • Dynamic generation of calendar events, database inserts, or API calls.

Vision (via Vision‑Enabled Models)

While Gemma 3 12B itself is a text‑only model, the Resayil platform also offers vision models (e.g., Qwen3‑VL 235B). You can chain a vision model’s output into Gemma 3 12B for multimodal pipelines.

Thinking Models & Tool Use

Resayil lists thinking models and tool use as features. By prompting Gemma 3 12B with chain‑of‑thought instructions, you can obtain more reasoning‑rich answers, and you can invoke external tools via function calling to extend its capabilities.

All of these capabilities are accessed through the same endpoints:

  • /v1/chat/completions – primary chat interface, supports streaming and function calling.
  • /v1/messages – message‑based API for more granular control.

Top Use Cases for Gemma 3 12B via the Resayil API {#use-cases}

  1. Arabic‑language chatbots – Build conversational agents that understand dialects, respond with culturally appropriate phrasing, and integrate with CRM systems via function calling.
  2. Multilingual customer support – Deploy a single backend that handles tickets in Arabic, English, French, and more, routing each language to specialized prompts.
  3. Code assistants with function calling – Use Gemma 3 12B to generate code snippets, then invoke a function that runs a linter or executes a test suite, returning results in the chat.
  4. Vision‑enabled pipelines – Combine a vision model (e.g., Qwen3‑VL 235B) to extract text from images, then feed that text to Gemma 3 12B for summarisation or translation.
  5. Content localisation – Generate marketing copy in Arabic from English source material, preserving brand voice while adapting cultural references.

Because Resayil charges per‑use credits, you can prototype these use cases without committing to a subscription, scaling cost linearly with traffic.

Getting API Access: Authentication, Endpoints, and Pricing {#api-access}

  1. Create an account on the Resayil portal (https://llm.resayil.io) and obtain an API key.
  2. Authentication – Include the key in the Authorization: Bearer <YOUR_KEY> header for every request.
  3. Key endpoints:
    • GET /v1/health – health check.
    • GET /v1/models – list all 40 models; look for gemma3:12b.
    • GET /v1/models/{id} – model details.
    • POST /v1/chat/completions – chat interface (streaming, function calling).
    • POST /v1/messages – message‑based API.
    • POST /v1/messages/count_tokens – token counting.
    • GET /v1/pricing – view credit pricing.
    • GET /v1/pricing/topups – purchase additional credits.
  4. Pricing – All usage is billed in USD credits. No subscription tiers are required; you simply purchase credits via Stripe or PayPal as needed.
  5. Top‑up – Call /v1/pricing/topups to add more credits to your balance.

The straightforward credit model makes budgeting transparent: each token consumed deducts a known amount of credit, which you can monitor via the portal dashboard.

Ready to try Resayil LLM API?

Start Free

Integration Examples: OpenAI SDK, LangChain, Python, and cURL {#integration-examples}

OpenAI SDK (Python)

import openai
import os

openai.api_key = os.getenv("RESAYIL_API_KEY")
openai.api_base = "https://llm.resayil.io/v1"

response = openai.ChatCompletion.create(
    model="gemma3:12b",
    messages=[{"role": "user", "content": "اكتب مقالة عن الذكاء الاصطناعي بالعربية"}],
    stream=False,
    temperature=0.7,
)
print(response.choices[0].message.content)

Streaming with Function Calling (OpenAI SDK)

import openai, os, json
openai.api_key = os.getenv("RESAYIL_API_KEY")
openai.api_base = "https://llm.resayil.io/v1"

functions = [{
    "name": "store_feedback",
    "description": "Save user feedback to a database",
    "parameters": {
        "type": "object",
        "properties": {
            "rating": {"type": "integer", "minimum": 1, "maximum": 5},
            "comment": {"type": "string"}
        },
        "required": ["rating"]
    }
}]

response = openai.ChatCompletion.create(
    model="gemma3:12b",
    messages=[{"role": "user", "content": "أعطيني تقييمًا للرد السابق"}],
    functions=functions,
    function_call="auto",
    stream=True,
)
for chunk in response:
    if "function_call" in chunk.choices[0].delta:
        print(json.dumps(chunk.choices[0].delta.function_call, ensure_ascii=False))

Python requests Example

import requests, os, json
url = "https://llm.resayil.io/v1/chat/completions"
headers = {
    "Authorization": f"Bearer {os.getenv('RESAYIL_API_KEY')}",
    "Content-Type": "application/json"
}
payload = {
    "model": "gemma3:12b",
    "messages": [{"role": "user", "content": "Translate this to Arabic: 'The future of AI is bright.'"}],
    "temperature": 0.5
}
resp = requests.post(url, headers=headers, json=payload)
print(resp.json())

cURL Example (Streaming)

curl https://llm.resayil.io/v1/chat/completions \
  -H "Authorization: Bearer $RESAYIL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
        "model": "gemma3:12b",
        "messages": [{"role": "user", "content": "اكتب قصة قصيرة عن طائر في الصحراء"}],
        "stream": true
      }'

LangChain Integration

from langchain.llms import OpenAI

llm = OpenAI(
    model_name="gemma3:12b",
    openai_api_key=os.getenv("RESAYIL_API_KEY"),
    openai_api_base="https://llm.resayil.io/v1",
)
print(llm.predict("ما هو تعريف التعلم الآلي؟"))

All examples use the OpenAI‑compatible endpoint, so you can swap the base URL to point at Resayil without changing code.

Pricing & Billing: Pay‑per‑Use Credits in USD {#pricing}

Resayil’s pricing is credit‑based and denominated in USD. You purchase credits through Stripe or PayPal; each token generated or processed consumes a fraction of a credit. There are no subscription tiers or hidden fees—only the cost of the tokens you actually use.

  • View rates – GET /v1/pricing returns the current credit‑per‑token price.
  • Top‑up credits – GET /v1/pricing/topups shows bulk purchase options for larger projects.
  • Monitor usage – The portal dashboard displays real‑time credit consumption, helping you stay within budget.

Compared with subscription‑based APIs that charge a fixed monthly fee regardless of usage, Resayil’s model offers maximum flexibility for developers who experience variable traffic or are experimenting with new ideas.

FAQ

Q: Is Gemma 3 12B available on the LLM Resayil Portal?

A: Yes. Gemma 3 12B is one of the 40 active models. You can confirm its presence by calling the /v1/models endpoint.

Q: Does the API support streaming responses for Gemma 3 12B?

A: Yes. Streaming is a verified feature. Include "stream": true in the /v1/chat/completions request to receive token‑by‑token output.

Q: Can I use Gemma 3 12B for Arabic text generation?

A: Absolutely. The portal explicitly supports Arabic language generation and multilingual use cases.

Q: What payment methods are accepted for using Gemma 3 12B?

A: Payments are accepted via Stripe and PayPal, and all billing is in USD.

Q: How do I integrate Gemma 3 12B with LangChain or n8n?

A: Resayil provides official integrations for LangChain, n8n, LiteLLM, and is compatible with the OpenAI and Anthropic SDKs. Set the openai_api_base to https://llm.resayil.io/v1 and use the model slug gemma3:12b in your workflow.

Call to Action

Ready to start building Arabic‑enabled AI experiences? Visit the Resayil pricing page to view credit rates, register for an API key, and explore the documentation for full integration guides.

Gemma 3 12B Resayil API
All Articles Read More Articles