Developers
Documentation
A technical guide to integrating leading AI models through Oblion's unified API.
On this page
Model Groups
API Group Type Comparison Guide
Group Types and Characteristics
| Group | Type | Rate | Supported Models |
|---|---|---|---|
| Default | Mixed ChatGPT (Azure channel) + Claude + MJ (fast model) + Domestic models (DeepSeek + Qwen) | Official rate x1 | OpenAI, Claude, Domestic models |
| Enterprise High-Availability | Domestic models (DeepSeek + Qwen) | Official rate x1 | Domestic models |
| Premium Grok | Grok models (select Grok models) | Official rate x5 | Grok |
| Premium Gemini | Gemini (Google channel) models | Official rate x1 | Gemini |
| Official-Relay Gemini | Gemini (Google channel) models, higher cost (more accounts, higher cost) | Official rate x3 | Gemini |
| Pure Azure | ChatGPT models only (Azure channel) + Domestic models (Doubao + DeepSeek) | Official rate x1.5 | OpenAI, Claude, Domestic models |
| Official-Relay | ChatGPT (Azure channel) + ChatGPT (official-relay channel) + Domestic models (Azure channel used first, falls back to official-relay when unavailable) | Official rate x3 | OpenAI, Domestic models |
| Official-Relay OpenAI | ChatGPT models (official-relay channel) + Azure channel (fallback channel) | Official rate x6 | OpenAI |
| Premium Official-Relay OpenAI | ChatGPT models (official-relay channel), higher cost (more accounts) | Official rate x8 | OpenAI |
| Limited-Time Trial | Supports GPT + Claude + Gemini + Grok | Official rate x1.4 | OpenAI, Claude |
| Limited-Time Special | Supports Domestic models (DeepSeek + Qwen) + Gemini (Google channel) models + ChatGPT (Azure channel) | Official rate x0.6 | Gemini, Domestic models |
| Official-Relay Claude 2 | Claude (AWS official-relay channel) | Official rate x6 | Claude |
| Official-Relay Claude 3 | Claude (AWS official-relay channel + Anthropic official-relay channel) | Official rate x12 | Claude |
| Direct Claude | Claude (Anthropic official-relay channel) | Official rate x16 | Claude |
| Claude Code Exclusive | Claude Code | Official rate x1.5 | Claude Code |
Pros and Cons Comparison
| Feature | Limited-Time Trial | Default (Mixed) | Azure | Official-Relay | Official-Relay / Premium Official-Relay OpenAI | Premium Gemini | Official-Relay Gemini | Official-Relay Claude | Direct Claude |
|---|---|---|---|---|---|---|---|---|---|
| Pros | Same as the official ChatGPT site, smarter, cheaper, but responses differ slightly from the API — likely a different backing version despite being the same model | Same as the official ChatGPT site, smarter, cheaper | Many channels, ample backup capacity, high volume, high concurrency, no review, supports FC/TC | High speed, high volume, high concurrency, no review, supports FC/TC | Same as the official ChatGPT site, high speed, high volume, high concurrency, more accounts, more stable | Same as the official site, smarter, cheaper, high volume, stable | Same as the official site, smarter, cheaper, high volume, more stable, more accounts | From official Amazon and Anthropic channels, many channels, ample backup capacity, high volume, high concurrency | From the official Anthropic channel, more accounts, more stable |
| Cons | Unofficial interface | Unofficial interface | Azure can occasionally lag | Affected by official API stability | Affected by official API stability | Affected by official API stability | Higher cost, higher rate multiplier, affected by official API stability | Affected by official API stability | Affected by official API stability |
| Details | Sourced entirely from OpenAI's official channel, subject to routine health checks | Mixed Azure + official-relay | Uses enterprise no-review Azure, supports FC/TC, supports 10M TPM, falls back to OpenAI on failure | All via the official OpenAI channel, supports FC/TC, supports 10M TPM, uses large OpenAI accounts, high concurrency | Via the official OpenAI channel, supports FC/TC, supports 10M TPM, uses large OpenAI accounts, high concurrency | Official Google API, large accounts, high concurrency | Official Google API, large accounts, high concurrency | All via official channels, supports function calling and Claude Code, large accounts, high concurrency | All via official channels, supports function calling and Claude Code, large accounts, high concurrency |
Model Support Matrix
| Model Type | Limited-Time Trial | Default (Mixed) | Azure | Official-Relay |
|---|---|---|---|---|
| GPT-4 | Fully supported | Fully supported | Azure, fully supported | OpenAI, fully supported |
| GPT-3.5 | Fully supported | Fully supported | Azure, fully supported | OpenAI, fully supported |
| Other base OpenAI models | Fully supported | Fully supported | Fully supported | Fully supported |
| Midjourney | Not supported | Fully supported | Not supported | Not supported |
| Domestic models | Not supported | Fully supported | Partially supported | Not supported |
| Claude | Fully supported | Fully supported | Fully supported | Fully supported |
Channel Explanations
- ChatGPT official-relay channel: sourced from openai.com
- Azure channel: sourced from Microsoft Azure
- Gemini Google channel: sourced from Google
- Domestic models: sourced from their respective official providers
- Claude AWS official-relay: sourced from Amazon
- Claude Anthropic official-relay: sourced from Anthropic
You can set the group for a token on the Tokens page.