Before your first top-up
No credit card required, and trial calls never expire.
Trial calls 3 total, all free models
Free AI Models
Free models are for experimenting, learning and shipping without a bill — AIHubMix covers the inference cost for all 14 models on this page, and every model answers all three protocols: Chat Completions, Messages and Responses. Every account starts with 3 trial calls, no credit card, and a $1 top-up switches you to permanent daily quotas.
After a one-time $1 top-upRecommended
Top up once — any amount from $1 — and every free model switches to daily quotas with no expiry, reset every day.
Requests 5/dayRequests 2/minTokens 10K/dayResets daily
Usage
69.9K
71
4 of 14
- ling-3.0-flash-fin-free
- gpt-4.1-free
- ling-3.0-flash-fin-free
- gpt-oss-20b-free
- gpt-4.1-free
- ling-2.6-flash-free
Which models that traffic went to
- Ling 3.0 Flash Fin (free)99.7%69.7K
- GPT 4.1 (free)0.3%195
- Ling 3.0 Flash Fin (free)93.0%66
- GPT Oss 20B (free)4.2%3
- GPT 4.1 (free)1.4%1
- Ling 2.6 Flash (free)1.4%1
All 14 Free Models
Open in model list| Model | Modalities | Context | Rate limit | Paid version |
|---|---|---|---|---|
 coding-glm-4.6-free | Takes text, returns text. | 200K | 2/day · 1/min | vs paid → |
 gemini-2.0-flash-free | — | — | 5/day · 2/min | vs paid → |
| gemini-3-flash-preview-free | Takes text, vision, audio, video, PDF. Output modality not published. | 1.05M | 5/day · 2/min | vs paid → |
| ?gemini-3.1-flash-image-preview-freeNew | — | — | 5/day · 2/min | vs paid → |
| gemini-3.8-flash-freeNew | Takes text, vision, audio, video, PDF. Output modality not published. | 1.05M | 5/day · 2/min | vs paid → |
| gemma-4-26b-a4b-it-free | Takes text, vision, video, returns text. | 262K | 5/day · 2/min | — |
| gemma-4-31b-it-free | Takes text, vision, video, returns text. | 262K | 5/day · 2/min | — |
 gpt-4.1-free | Takes text, vision, PDF, returns text. | 1.05M | 5/day · 2/min | vs paid → |
| ?gpt-image-2-freeNew | Takes text, vision. Output modality not published. | — | 5/day · 2/min | vs paid → |
| gpt-oss-20b-free | Takes text, returns text. | 131K | 5/day · 2/min | vs paid → |
 ling-2.6-flash-free | Takes text, returns text. | 262K | 5/day · 2/min | — |
 ling-3.0-flash-fin-freeNew | Takes text. Output modality not published. | 262K | 5/day · 2/min | — |
 nemotron-3.5-content-safety-free | Takes text, vision, returns text. | 131K | 5/day · 2/min | — |
 qwen3.6-plus-preview-free | Takes text, returns text. | 1M | 5/day · 2/min | vs paid → |
How to Start
- 1
Create an account
Email or OAuth, no credit card required. 3 trial calls are available immediately.
- 2
Create an API key
One key works for every model on the platform, free and paid alike. Open the API keys page →
- 3
Call a free model
Point any OpenAI-compatible client — Cursor, Cline, Cherry Studio, LiteLLM — at the gateway and pick a
-freemodel ID.
POST /v1/chat/completions
curl https://aihubmix.com/v1/chat/completions \
-H "Authorization: Bearer sk-***" \
-H "Content-Type: application/json" \
-d '{
"model": "ling-3.0-flash-fin-free",
"messages": [{"role":"user","content":"Hi"}]
}' # gpt-oss-20b-free / gpt-4.1-free
A -free ID like gpt-oss-20b-free — the most-called free model this month — is the same upstream model as gpt-oss-20b, served as a subsidized free variant. Swap the suffix off when you need paid throughput; nothing else in your code changes.
Free Models on AIHubMix
Are these models really free?
Yes. All 14 models bill $0 for input and output tokens. AIHubMix subsidizes the inference cost; usage is bounded by quotas instead of a bill.
What do I get before topping up?
3 trial calls, shared across all free models, with no expiry. When they run out the API keeps answering with a note that the trial is used up — your integration never sees a hard error.
What do I get after topping up?
Any top-up of $1 or more permanently switches your account to daily quotas: 2 requests per minute, 5 requests per day and 10K tokens per day, shared across the free catalog and reset every day. Models with a higher request weight count as several requests per call — the exact caps are in the table. Past a cap the API returns 429 until the next minute or day starts.
What is the difference between GPT Oss 20B (free) and gpt-oss-20b?
Price is the only difference — they are the same upstream model with identical capability. The -free ID bills $0 within the quotas above; the paid ID bills per token with no daily cap. Switching between them is a one-line model-name change.
Keep Exploring
All Models
The full catalog — 750 models across chat, image, video and audio, with per-token pricing.
Browse all models →LLM Router
Fill in auto as the model ID and the gateway routes each request to the optimal model by cost and quality.
Apps
Ready-to-use apps in the AIHubMix app store — and model discounts beyond the free catalog.
Browse the app store →