API Reference
Access OpenAI, Anthropic, and Google Gemini models through one gateway, using the SDKs you already know. Change two lines — the base URL and your key — and you're live.
Quickstart
Install the OpenAI SDK, point it at the AICreditMart endpoint, and use your API key.
from openai import OpenAI client = OpenAI( base_url="https://api.aicreditmart.com/v1/", api_key="YOUR_API_KEY", ) response = client.chat.completions.create( model="gpt-5.4", messages=[{"role": "user", "content": "Hello!"}], ) print(response.choices[0].message.content)
base_url and api_key.Authentication
Pass your API key as the api_key in the SDK client. It's sent as a standard Bearer token and tracks your usage and billing.
api_key = "sk-bf-xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
Keep your key private. If a key is exposed, contact us to rotate it.
Base URLs
Three endpoints, depending on which SDK you use:
https://api.aicreditmart.com/v1/https://api.aicreditmart.com/anthropichttps://api.aicreditmart.com/genaiUse the OpenAI base URL for all OpenAI models, the Anthropic base URL with the anthropic SDK for Claude models, and the Gemini base URL with Google's google-genai SDK for Gemini models.
Rate Limits
Rate limits are generous and set per API key. Most workloads run without hitting them. If you expect high sustained throughput or need a dedicated limit, contact us and we'll raise it for your key.
Regions
Models run primarily on US-based infrastructure (Azure US, Google US, and others). If you have data-residency requirements, EU region routing is available on request — contact us to enable it for your key.
Tracking Usage
Every account has a dedicated dashboard with real-time usage and cost tracking, broken down by model, tokens, and spend — so you always know exactly what you're using.
Sign in to view your usage at app.aicreditmart.com.
Anthropic Models
Use Anthropic's anthropic SDK pointed at the Anthropic base URL. Send the Model ID exactly as shown.
Text
| Model ID | Notes |
|---|---|
claude-sonnet-5 | Latest Sonnet — top-tier intelligence at Sonnet pricing |
claude-opus-4-8 | Most capable Claude for complex challenges |
claude-opus-4-7 | Previous Opus generation, highly capable |
claude-opus-4-6 | Earlier Opus generation |
claude-sonnet-4-6 | Fast, balanced model for everyday tasks |
claude-haiku-4-5 | Fastest, most affordable Claude for high-volume tasks |
Example — Text (claude-sonnet-4-6)
import anthropic client = anthropic.Anthropic( base_url="https://api.aicreditmart.com/anthropic", api_key="YOUR_API_KEY", # your sk-bf virtual key, not a real Anthropic key ) message = client.messages.create( model="claude-sonnet-4-6", max_tokens=1000, messages=[ {"role": "user", "content": "Tell me about Stockholm in one sentence."} ], ) print(message.content[0].text)
OpenAI Models
Send the Model ID exactly as shown.
Text & reasoning
| Model ID | Notes |
|---|---|
gpt-5.6-sol | Most capable 5.6 model for complex reasoning |
gpt-5.6-terra | Balanced 5.6 model, strong price-performance |
gpt-5.6-luna | Fastest, most cost-efficient 5.6 model |
gpt-5.5 | Frontier model with configurable reasoning |
gpt-5.4 | Affordable model for coding and professional work |
gpt-5.4-mini | Strongest mini for coding and computer use |
gpt-5.4-nano | Cheapest 5.4-class model for high-volume tasks |
gpt-5.2 | Previous frontier model, configurable reasoning |
gpt-5.1 | Strong coding and agentic tasks, configurable reasoning |
gpt-5 | Reasoning model for coding and agentic tasks |
gpt-5-mini | Near-frontier, low-latency, high-volume workloads |
gpt-5-nano | Fastest, most cost-efficient GPT-5 |
gpt-4.1 | Smartest non-reasoning model |
gpt-4.1-mini | Smaller, faster GPT-4.1 |
gpt-4.1-nano | Fastest, most cost-efficient GPT-4.1 |
gpt-4o | Fast, intelligent, flexible multimodal model |
gpt-4o-mini | Fast, affordable model for focused tasks |
Image
| Model ID | Notes |
|---|---|
gpt-image-2 | State-of-the-art image generation |
gpt-image-1.5 | Previous-generation image model |
gpt-image-1-mini | Cost-efficient image generation |
Video
| Model ID | Notes |
|---|---|
sora-2 | Flagship video generation with synced audio |
Embeddings
| Model ID | Notes |
|---|---|
text-embedding-3-large | Most capable embedding model |
text-embedding-3-small | Small, fast embedding model |
Example — Text (gpt-5.4)
from openai import OpenAI client = OpenAI( base_url="https://api.aicreditmart.com/v1/", api_key="YOUR_API_KEY", ) response = client.chat.completions.create( model="gpt-5.4", messages=[{"role": "user", "content": "How far is New York from London?"}], ) print(response.choices[0].message.content) print(f"Tokens used: {response.usage.total_tokens}")
Example — Image (gpt-image-2)
from openai import OpenAI import base64 client = OpenAI( base_url="https://api.aicreditmart.com/v1/", api_key="YOUR_API_KEY", ) response = client.images.generate( model="gpt-image-2", prompt="A majestic horse standing in a field", n=1, size="1024x1024", quality="medium", output_format="png", ) if response.data[0].b64_json: img = base64.b64decode(response.data[0].b64_json) with open("horse.png", "wb") as f: f.write(img) print("Saved horse.png")
Example — Video (sora-2)
import time from openai import OpenAI client = OpenAI( base_url="https://api.aicreditmart.com/v1/", api_key="YOUR_API_KEY", ) # 1. Create the video video = client.videos.create( model="sora-2", prompt="A cat playing piano in a jazz bar", size="1280x720", seconds="4", ) # 2. Poll until complete while video.status in ["queued", "in_progress"]: time.sleep(5) video = client.videos.retrieve(video.id) print(f"Status: {video.status}") # 3. Download if video.status == "completed": content = client.videos.download_content(video.id, variant="video") content.write_to_file("output.mp4") print("Saved output.mp4")
Example — Embeddings (text-embedding-3-small)
from openai import OpenAI client = OpenAI( base_url="https://api.aicreditmart.com/v1/", api_key="YOUR_API_KEY", ) response = client.embeddings.create( model="text-embedding-3-small", input="The quick brown fox jumps over the lazy dog.", ) print(f"Dimensions: {len(response.data[0].embedding)}")
Google Gemini Models
Use Google's google-genai SDK pointed at the Gemini base URL. Send model names with the models/ prefix.
Text
| Model ID | Notes |
|---|---|
gemini-3.5-flash | Most intelligent flash for agentic and coding |
gemini-3.5-flash-lite | Fast, low-cost model for high-volume tasks |
gemini-3.1-pro-preview | Advanced reasoning and agentic coding (preview) |
gemini-3.1-flash-lite | Frontier-class performance at low cost |
gemini-3-flash-preview | Frontier-class performance, low cost (preview) |
gemini-2.5-pro | Advanced reasoning and coding for complex tasks |
gemini-2.5-flash | Best price-performance for high-volume reasoning |
gemini-2.5-flash-lite | Fastest, most budget-friendly 2.5 model |
Image
| Model ID | Notes |
|---|---|
gemini-3-pro-image | Studio-quality 4K image generation and editing |
gemini-3.1-flash-image | High-efficiency image generation, optimized for speed |
gemini-2.5-flash-image | Native image generation for fast creative workflows |
Example — Text (gemini-3.5-flash)
from google import genai from google.genai import types client = genai.Client( api_key="YOUR_API_KEY", http_options=types.HttpOptions(base_url="https://api.aicreditmart.com/genai"), ) response = client.models.generate_content( model="models/gemini-3.5-flash", contents="Tell me about London in one sentence.", ) print(response.text)
Example — Image (gemini-3.1-flash-image)
from google import genai from google.genai import types client = genai.Client( api_key="YOUR_API_KEY", http_options=types.HttpOptions(base_url="https://api.aicreditmart.com/genai"), ) MODEL = "gemini-3.1-flash-image" r = client.models.generate_content( model=f"models/{MODEL}", contents="A photorealistic golden retriever wearing a red cape, studio lighting", config=types.GenerateContentConfig(response_modalities=["IMAGE"]), ) for part in r.candidates[0].content.parts: if part.inline_data and part.inline_data.data: with open(f"{MODEL}.png", "wb") as f: f.write(part.inline_data.data) print(f"Saved {MODEL}.png")
image_config=types.ImageConfig(image_size="4K") inside the config for up to 4K (on gemini-3.1-flash-image and gemini-3-pro-image).