Pay in rupees · No international card needed
Top AI modelsfrom ₹100.Pay with UPI
One API key for models from OpenAI, Anthropic, Google, xAI and Alibaba. Top up with UPI or card, pay up to 90% below list price, and your credit never expires.
₹200 free credit for new accounts, added when you verify your email
https://www.maxonforge.com/v1 · /v1/messages · /v1/models
Price per 1M tokens
list · you pay
Anthropic
Claude Fable 5.1
- Input
- $10.00
- $1.00
- Output
- $50.00
- $5.00
OpenAI
GPT-6 Astra
- Input
- $10.00
- $1.00
- Output
- $50.00
- $5.00
USD per 1M tokens. List price is the provider's published rate; rates can change. All live rates
Why the price is up to 90% lowerWe buy in bulk and pass the wholesale rate through
No magic and no catch: we buy inference from suppliers in bulk, keep no reseller markup, and pass the wholesale rate straight through. The catalogue prints their list price next to ours, so you can check the saving before you pay.
90% below the published list price
Every model is billed at one tenth of its verified reference rate. Claude Fable 5.1 at $1 in / $5 out per 1M tokens instead of $10 / $50 — the same model, the same API, one tenth of the price.
We buy inference in bulk
We commit supplier capacity upfront and pay wholesale, then pass that rate straight through. No per-seat markup, no agency fee, no reseller chain adding its margin on top of yours.
Limited-time launch pricing
The 90% discount is a launch offer while we grow committed volume. There is no coupon code and nothing to unlock: the price you see in the catalogue is the price you are charged.
Official price and our price, side by side
The catalogue prints the provider's list price next to ours on the same row, so the saving is checkable before you spend a rupee — not a claim you have to trust.
No subscription, no expiry
Top up from ₹100 and spend it across every supported model. Credits never expire and each request is metered at the published per-token rate, so you pay only for what you actually use.
A saving you can calculate
Spend $10 here and you get roughly $100 of usage at list prices. Your dashboard shows the official value of your balance, so the saving is a number rather than a slogan.
One prepaid balance for every modelTop up from ₹100. It never expires
Pay once. Receive a defined amount of AI credit — your top-up plus its server-applied bonus. Each request spends it at the published model rate, and the balance never expires.
INR uses MaxonForge's fixed display rate ($1 = ₹100) — a display rate, not a market exchange rate. Checkout always shows the exact price before you pay.
Minimum top-up ₹100 at checkout. Larger top-ups earn a tier bonus. The packs below are examples of common amounts.
Starter
Common top-up amount
Minimum top-up · prepaid, never expires
- $1.00 in AI usage
- Published customer model rates
- Works with the API and usage dashboard
- Credits never expire
- No monthly subscription
Growth
Common top-up amount
One-time purchase · prepaid, never expires
- $5.10 in AI usage
- Published customer model rates
- Works with the API and usage dashboard
- Credits never expire
- No monthly subscription
Power
Common top-up amount
One-time purchase · prepaid, never expires
- $21.00 in AI usage
- Published customer model rates
- Works with the API and usage dashboard
- Credits never expire
- No monthly subscription
Scale
Common top-up amount
One-time purchase · prepaid, never expires
- $110.00 in AI usage
- Published customer model rates
- Works with the API and usage dashboard
- Credits never expire
- No monthly subscription
Where the 10× value comes from: the customer price is 90% below the verified reference price, so $1 of AI credit buys about $10 of usage at the providers' own prices. That is a pricing relationship — it is never added again as a bonus on top of your balance.
Actual usage is metered per request using the selected model's configured pricing — the balance decreases by exactly what each request consumes, shown per request in your dashboard. No token estimates, no rounding games.
Credit never expires. You are only ever charged for the tokens you use, at the published rate.
Before you pay. The exact MaxonForge price is shown at checkout — no provider surcharges are added to it.
Spend $10, use about $100 of AIThe discount is already in every model rate
Pick a top-up and see what the same usage would cost at the providers' published list prices. No coupon, no code — the discount is already in every model rate.
You pay here
$10
one-time · never expires
Same usage at list price
$100
what other platforms charge
You keep
$90
saving at 90% below list
Comparison uses our published launch discount of 90% below verified reference prices. Discounts vary per model; exact rates are on the catalogue page. Limited-time launch pricing — rates can change, and the catalogue is always the current price.
Model catalog
Model IDs, providers, and published rates from the live platform catalog. Pick a model that fits the task — one API key covers every model you enable.
- Model ID
gpt-6-astraMaxonForge input price$1.00 per 1M input tokensMaxonForge output price$5.00 per 1M output tokensOfficial input price$10.00 per 1M input tokens (provider list price)Official output price$50.00 per 1M output tokens (provider list price)Context window—Promotion90% offActual usage is metered per request at these rates — the price shown here is the price the billing engine charges, including any active promotion. One API key covers every model you enable.
- Model ID
claude-fable-5-1MaxonForge input price$1.00 per 1M input tokensMaxonForge output price$5.00 per 1M output tokensOfficial input price$10.00 per 1M input tokens (provider list price)Official output price$50.00 per 1M output tokens (provider list price)Context window1M tokensPromotion90% offActual usage is metered per request at these rates — the price shown here is the price the billing engine charges, including any active promotion. One API key covers every model you enable.
- Model ID
claude-fable-5MaxonForge input price$1.00 per 1M input tokensMaxonForge output price$5.00 per 1M output tokensOfficial input price$10.00 per 1M input tokens (provider list price)Official output price$50.00 per 1M output tokens (provider list price)Context window1M tokensPromotion90% offActual usage is metered per request at these rates — the price shown here is the price the billing engine charges, including any active promotion. One API key covers every model you enable.
- Model ID
claude-opus-5-5MaxonForge input price$0.400 per 1M input tokensMaxonForge output price$2.00 per 1M output tokensOfficial input price$4.00 per 1M input tokens (provider list price)Official output price$20.00 per 1M output tokens (provider list price)Context window1M tokensPromotion90% offActual usage is metered per request at these rates — the price shown here is the price the billing engine charges, including any active promotion. One API key covers every model you enable.
- Model ID
claude-opus-5MaxonForge input price$0.500 per 1M input tokensMaxonForge output price$2.50 per 1M output tokensOfficial input price$5.00 per 1M input tokens (provider list price)Official output price$25.00 per 1M output tokens (provider list price)Context window1M tokensPromotion90% offActual usage is metered per request at these rates — the price shown here is the price the billing engine charges, including any active promotion. One API key covers every model you enable.
- Model ID
claude-sonnet-5-5MaxonForge input price$0.240 per 1M input tokensMaxonForge output price$1.20 per 1M output tokensOfficial input price$2.00 per 1M input tokens (provider list price)Official output price$10.00 per 1M output tokens (provider list price)Context window1M tokensPromotion88% offActual usage is metered per request at these rates — the price shown here is the price the billing engine charges, including any active promotion. One API key covers every model you enable.
- Model ID
claude-sonnet-5MaxonForge input price$0.240 per 1M input tokensMaxonForge output price$1.20 per 1M output tokensOfficial input price$2.00 per 1M input tokens (provider list price)Official output price$10.00 per 1M output tokens (provider list price)Context window1M tokensPromotion88% offActual usage is metered per request at these rates — the price shown here is the price the billing engine charges, including any active promotion. One API key covers every model you enable.
- Model ID
claude-haiku-5-5MaxonForge input price$0.240 per 1M input tokensMaxonForge output price$1.20 per 1M output tokensOfficial input price$1.00 per 1M input tokens (provider list price)Official output price$5.00 per 1M output tokens (provider list price)Context window200K tokensPromotion76% offActual usage is metered per request at these rates — the price shown here is the price the billing engine charges, including any active promotion. One API key covers every model you enable.
- Model ID
claude-opus-4-8MaxonForge input price$0.500 per 1M input tokensMaxonForge output price$2.50 per 1M output tokensOfficial input price$5.00 per 1M input tokens (provider list price)Official output price$25.00 per 1M output tokens (provider list price)Context window1M tokensPromotion90% offActual usage is metered per request at these rates — the price shown here is the price the billing engine charges, including any active promotion. One API key covers every model you enable.
- Model ID
claude-opus-4-7MaxonForge input price$0.500 per 1M input tokensMaxonForge output price$2.50 per 1M output tokensOfficial input price$5.00 per 1M input tokens (provider list price)Official output price$25.00 per 1M output tokens (provider list price)Context window1M tokensPromotion90% offActual usage is metered per request at these rates — the price shown here is the price the billing engine charges, including any active promotion. One API key covers every model you enable.
Usage · per request
example
| Model | Tokens in / out | Cost |
|---|---|---|
| claude-fable-5-114:02:11 | 12,480 / 1,920 | $0.0221 |
| gpt-6-astra14:02:37 | 3,210 / 860 | $0.0075 |
| claude-fable-5-114:03:05 | 48,900 / 4,150 | $0.0697 |
| gpt-6-astra14:04:18 | 920 / 310 | $0.0025 |
See what every request costsMetered per request, shown in your dashboard
Prepay what you need, choose a supported model, and see what each request consumes. A clear balance makes it easier to plan your next project.
Connect your first client.
Follow the documented setup for your client. Compatibility depends on your key scope, base URL, and selected model.
Read the setup guide- 01
Swap the base URL
Point your client at https://www.maxonforge.com/v1 — it serves both OpenAI-compatible and Anthropic Messages formats.
- 02
Choose a supported model
Copy a model ID from the catalog and check that your key has the required scope.
- 03
Send a first request
Follow the quick start, then review requests, tokens, and balance in your dashboard.
Works where you already buildCoding agents, automation, writing and apps
Examples of what the platform supports — through the API and the published model catalog. Your balance and usage work the same in every one of them.
Coding with agents
Point coding agents and CLI tools at the OpenAI-compatible base URL. Use supported models for reviews, refactors, and test runs.
Automation & agents
Connect workflow tools like n8n to supported models with your own key. Scope the key to the models each automation needs.
Writing & research
Draft, rewrite, and summarize through the API from the tool you already use. Swap the model id when a task calls for a different strength.
Cost tracking
Watch request-level usage and remaining balance in the dashboard, then use the published rates to plan the next top-up.
Application integration
Swap the base URL in an existing OpenAI or Anthropic SDK call. Streaming is supported on both wire formats.
Model experimentation
Compare published rates and context windows in the catalog, then try the same prompt across models to see what fits.
API, catalogue, docs and dashboardEverything you need on one account
Use the unified API, check live models and rates in the catalogue, read the setup docs, and track every request in the dashboard.
Unified API
OpenAI- and Anthropic-compatible endpoints. One API key, every enabled model. Drop into existing code with a base-URL change.
- OpenAI SDK compatible
- Anthropic SDK compatible
- Streaming responses
- Function calling support
Model catalogue
Check every published model, its context window, and the live customer rate before you spend. Official provider list prices sit beside the customer rate.
- Live published rates
- Official provider list prices
- Availability status per model
- Search and compare
Documentation
Quick start, authentication, OpenAI-compatible and Anthropic Messages examples, plus setup guides for the clients you already use.
- Quick-start guide
- OpenAI & Anthropic examples
- Client setup guides
- API key status checker
Dashboard
Manage API keys, track usage per request, and review your balance. Detailed breakdowns by model, endpoint, and date.
- Per-request usage tracking
- API key management
- Export usage as CSV
- Billing history
Bring the client you already use.
Step-by-step setup for Claude Code, Cursor, Cline, n8n and the OpenAI SDK. Any other OpenAI-compatible client works with the same base URL and key.
Questions,answered plainly
Check the documentation for setup, key scopes, and billing details. You can also open a support ticket for account or payment questions.
MaxonForge is a platform for accessing multiple AI models through one account: an OpenAI- and Anthropic-compatible API, a published model catalogue with live rates, and a usage dashboard. You fund the account with prepaid credit, then spend it at the published model rates. There is no subscription and nothing renews automatically.
You pay once, with a minimum top-up of ₹100, and the server converts your top-up into AI credit in USD at the published display rate. Each request spends that value at the model's published rate and is metered individually. Larger top-ups earn a server-applied bonus: +2% from ₹500, +5% from ₹1,000, +7% from ₹5,000, +10% from ₹10,000. Separately, the customer price being 90% below the verified reference price is what gives you roughly 10× the usage per dollar.
No. MaxonForge is pay-as-you-go only: you top up a balance and it is spent on the tokens you actually use. There is no monthly fee, no seat fee, no minimum spend, and nothing renews automatically. Credit you have already bought never expires.
The published catalog at /models (and GET /v1/models) lists every currently enabled model with its ID, context window and live rates. Availability follows the live routing configuration, so it can change — always check the catalog rather than a cached list.
Actual usage is metered per request using the selected model's configured pricing — input and output tokens are each priced per million, and the dashboard logs every request with what it consumed so you can see exactly what each call used.
Yes. Point any OpenAI-compatible client at https://www.maxonforge.com/v1 — it also serves Anthropic Messages at /v1/messages. One key covers every enabled model unless you restrict it to an allow-list, and streaming is supported on both wire formats.
MaxonForge routes each request to an enabled provider route configured for the requested model. Your key's scope and the model catalog determine what you can call — you never manage individual upstream provider accounts yourself.
If no route can serve the model right now, the request fails with a clear capacity or upstream error and nothing is billed for it. You can retry or switch to another model from the catalog.
No. Your prepaid balance stays on your account indefinitely. Use it at your own pace through the API and track it in the usage dashboard.
Yes. Keys default to every enabled model; you can narrow or reopen a key's model allow-list from the dashboard at any time, and the live catalogue shows which models are available before you call one.
That is what the API is for. Create a key, set the base URL, and ship — usage is metered per request and visible in your dashboard. Keep your key secret and scope its model access to what your application needs.
Checkout always shows the exact MaxonForge price before you pay — no provider surcharges are added on top. Online payments are verified by the payment gateway, and usage is added automatically the moment verification completes. Your AI usage balance is always shown in USD; where the payment provider settles in INR, a fixed display rate is used.
A change of mind is not a refund reason: once a top-up is credited to your balance, it is not refundable. Billing errors — a failed payment that was still charged, a duplicate charge, or a balance that is missing or short — are always corrected. If something we are responsible for breaks and we cannot fix it, we refund the affected amount. See the refund policy for full terms, or open a support ticket from your dashboard.
API · Dashboard · Model catalog
Start with ₹100 and one API key
Create your MaxonForge account, top up from ₹100, and start with the OpenAI- or Anthropic-compatible API or the usage dashboard.
₹200 free credit for new accounts — added when you verify your email
Minimum top-up ₹100 · Credits never expire · Exact price shown at checkout