AI access for apps that need to keep working.
SWForce AI API gives your software one simple endpoint for privately hosted AI models, prepaid credits, automatic recharge, customer-specific API keys, and controlled rate limits.
Use it when you want the convenience of an OpenAI-style API without sending every workflow through a public consumer account, sharing one company key across every app, or manually watching balances.
https://api.swforce.com/v1
One controlled API layer for real business systems.
The goal is not another chat window. It is a reliable AI layer that software can call repeatedly, track clearly, and pay for predictably.
Route models without rewrites
Point apps at one SWForce endpoint while we route traffic to the right hosted model behind it.
Separate keys per customer
Each account gets isolated keys, balances, usage history, and rate limits instead of one shared token.
Keep integrations online
Required auto-recharge reduces the chance of a production workflow stopping because credits ran out.
Start small, raise limits
Begin with conservative defaults, then increase request and token limits when usage justifies it.
Three steps from signup to first request.
The signup flow is intentionally simple: account, payment, API key. After that, most OpenAI-compatible apps only need a new base URL and your SWForce key.
Create account and add creditsCreate account
Enter company details, choose a plan, and accept auto-recharge terms.
Add credits
Use Stripe Checkout to fund the account and save the payment method.
Create API key
Generate a key, copy it once, and start calling the SWForce endpoint.
Use it from the tools your team already has.
Because the API follows the OpenAI request shape, it can fit into scripts, web apps, automations, support tools, and agent workflows with minimal changes.
Chat completion request
Send the same style of request many OpenAI-compatible clients already understand.
curl https://api.swforce.com/v1/chat/completions \
-H "Authorization: Bearer $SWFORCE_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"swforce/qwen3-32b-standard","messages":[{"role":"user","content":"Summarize this work order."}]}'
Good first use cases
- Summarize service tickets, emails, PDFs, and field notes.
- Classify inbound requests and route them to the right workflow.
- Draft responses using company-specific context and tone.
- Power internal agents without exposing backend model servers.
Start controlled, scale when usage proves it.
Every customer gets isolated API keys, their own LiteLLM team, and a prepaid balance with required auto-recharge. Higher limits can be approved for heavier production workloads.
General Customer
Default paid customer tier with generous production limits.
- 600 requests per minute
- 5,000,000 tokens per minute
- 25 parallel requests
- Prepaid credits with required auto-recharge
Premium Customer
Higher-throughput paid customer tier.
- 2,000 requests per minute
- 10,000,000 tokens per minute
- 50 parallel requests
- Prepaid credits with required auto-recharge