Pricing structure on the Ahur platform

The Ahur platform offers three different pricing models to cover three different kinds of need. Each one has its own advantages and its own best-fit use cases.

1. Ahur's own models

The AI models built in-house by Ahur are priced on token consumption. They're a good fit for anyone looking for strong value for money.

Ahur model pricing

def calculate_ahur_cost(input_tokens, output_tokens, model="ahur-base"):
    pricing = {
        "ahur-mini": (0.0005, 0.001),
        "ahur-base": (0.001, 0.002),
        "ahur-pro": (0.002, 0.004),
        "ahur-enterprise": (0.005, 0.010)
    }
    
    input_cost = (input_tokens / 1000) * pricing[model][0]
    output_cost = (output_tokens / 1000) * pricing[model][1]
    return input_cost + output_cost

# a worked example
cost = calculate_ahur_cost(1000, 500, "ahur-pro")
print(f"Total cost: {cost} Toman")

Advantages of the Ahur models

• Competitive pricing: cheaper than comparable models from abroad
• Persian support: tuned for the Persian language and culture
• Privacy: your data is never processed outside the country
• Local support: a Persian-speaking support team

2. Free models with limits

If you're just getting started or working on something small, there are free models available.

Free plans

Limits on the free models

• Maximum response length: 500 tokens per request
• Limited API access: suitable for testing and development
• Lower priority: requests are processed with a delay during peak times
• No advanced capabilities: no file analysis, no long-form summarization, and so on

3. Third-party models at list price

For access to the well-known global models, the price is exactly what the original provider charges.

Third-party model pricing

Why use third-party models through Ahur

• One unified bill: all of your usage on a single invoice
• Full Persian support: Persian interface and documentation
• Easy access: one platform for every model
• Local payment: pay from inside Iran

A full model comparison

Comparison table

Which model should you pick?

- Use the free models to try things out
- Start with Ahur Lite
- Move to a paid model once you have a feel for it

- The Ahur models perform best here
- Ahur Pro is ideal for professional workloads
- Lower cost and higher quality than models from abroad

- Claude-3 for complex analysis
- GPT-4 for specialized content generation
- Combine models for better results

- Ahur Base for generating Persian code
- GPT-3.5 for English documentation
- DeepSeek for cost-sensitive workloads

Things worth knowing

Optimizing your costs

1. Pick the right model: choose the most appropriate model for each job
2. Manage your tokens: keep prompts concise and efficient to bring costs down
3. Cache responses: use a cache for repeated requests
4. Monitor consumption: review your usage regularly

Cost calculator

function calculateTotalCost(model, inputTokens, outputTokens, monthlyUsage) {
    const pricing = {
        'ahur-base': { input: 0.001, output: 0.002 },
        'ahur-pro': { input: 0.002, output: 0.004 },
        'gpt-4': { input: 0.03, output: 0.06 },
        'claude-3': { input: 0.015, output: 0.075 }
    };
    
    const costPerRequest = (inputTokens / 1000) * pricing[model].input + 
                           (outputTokens / 1000) * pricing[model].output;
    
    return costPerRequest * monthlyUsage;
}

// example: working out the monthly cost
monthlyCost = calculateTotalCost('ahur-pro', 1000, 500, 100);
console.log(`Monthly cost: ${monthlyCost} Toman`);

Frequently asked questions

Can I switch between models?

Yes. You can switch between models at any time. Costs are calculated from your actual consumption on each model.

Are there any extra charges?

No. Charges are based purely on token consumption. There are no hidden fees and no mandatory monthly subscription.

Do I need a VPN for the third-party models?

No. Through the Ahur platform you can reach every third-party model without a VPN.

To start using the Ahur models and see exact rates, head to the pricing page.