API Documentation
llmair is OpenAI-compatible. If you can use the OpenAI API, you can use llmair — just change the base URL to https://api.llmair.ai/v1.
Get Your API Key
Sign up for free and get 100,000 tokens. No credit card required.
Base URL
https://api.llmair.ai/v1OpenAI-compatible endpoint — same format, just swap the URLAuthentication
Pass your API key in the Authorization header:
Authorization: Bearer $LLMAIR_API_KEYCode Examples
Make your first API call in under a minute.
curl https://api.llmair.ai/v1/chat/completions \
-H "Authorization: Bearer $LLMAIR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek-ai/DeepSeek-V3",
"messages": [{"role": "user", "content": "Explain quantum computing in one sentence."}],
"max_tokens": 100
}'Available Models
82 models from DeepSeek, Qwen, GLM, Kimi, and more. Prices shown per 1M tokens.
Flagship multimodal model
High-capability GPT-4
gpt-4oOpenAI-compatible router id
gpt-4-turboOpenAI-compatible router id
gpt-4-turbo-previewPreview Turbo variant
Fast, lightweight GPT-4o
Fast, older generation
Latest flagship, complex reasoning, code, long context
gpt-5.5OpenAI-compatible router id
Deep reasoning model
Frontier reasoning model
o1Reasoning model, router id
o1-previewReasoning preview
o1-miniLightweight reasoning
o3Frontier reasoning, router id
o3-miniCompact reasoning
Deep analysis, long-form writing
Snapshot Opus version
Balanced, thoughtful, capable
Fast and lightweight
Fast and lightweight
claude-3-opusOpenRouter-style id
claude-sonnet-4Latest Sonnet
claude-sonnet-4-20250514Latest Sonnet snapshot
Clear, balanced answers
Deepest Google model
gemini-2.5-proLatest Gemini Pro tier
gemini-2.5-flashFast, lightweight Gemini
Witty, slightly rebellious
Open-weights flagship
Mid-tier open model
Alt id for Llama 3 8B
Alt id for Llama 3 70B
Clear, structured European model
Direct chat model
Chat form of R1 reasoning
Fast, economical Qwen
Balanced Qwen tier
DeepSeek-V3Complex reasoning, research
DeepSeek-V3.1-TerminusComplex reasoning
DeepSeek-V3.2Complex reasoning, latest
DeepSeek-R1Advanced reasoning
DeepSeek-V4-FlashFast, cost-efficient
DeepSeek-V4-ProBalanced performance
Qwen3-8BFast, low-cost tasks
Qwen3-14BBalanced general purpose
Qwen3-32BStrong reasoning, coding
Qwen3-30B-A3B-Instruct-2507Efficient high-capability MoE
Qwen3-Coder-30B-A3B-InstructCode generation, debugging
Qwen3.5-27BGeneral purpose, long context
Qwen3.5-35B-A3BHigh capability, efficient
Qwen3.5-9BFast, low-cost tasks
Qwen3.5-4BMinimal cost, quick tasks
Qwen3.5-397B-A17BMaximum capability, research
Qwen3.6-27BLatest generation, reasoning
Qwen3-72B-InstructHigh capability, complex tasks
Qwen2.5-72B-Instruct-128KLong documents, massive context
GLM-4.5-AirLong context, Chinese language
GLM-4.7Advanced reasoning, long context
GLM-5Latest GLM, highest capability
GLM-5.1Latest version, research
GLM-Z1-9B-0414Fast reasoning, low cost
Hunyuan-A13B-InstructGeneral purpose, Tencent ecosystem
Hunyuan-MT-7BTranslation, multilingual
Step-3.5-FlashFast, cost-effective tasks
Nex-N2-ProAdvanced reasoning, agentic tasks
Seed-OSS-36B-InstructHigh capability, creative tasks
MiniMax-M2.5Long context, multimodal
deepseek-aiMaximum reasoning capability
deepseek-aiPremium DeepSeek performance
moonshotaiVery long context, research
moonshotaiLatest Kimi, longest context
zai-orgPremium GLM capability
QwenEfficient Pro tier
Qwen3-VL-8B-InstructVision + text, image understanding
Qwen3-VL-32B-InstructHigh-capability vision
Qwen3-VL-30B-A3B-InstructVision MoE, efficient
GLM-4.5VVision + text, Chinese focus
KolorsImage generation, creative
Qwen-ImageImage understanding
SenseVoiceSmallSpeech recognition, voice AI
CosyVoice2-0.5BVoice synthesis, TTS
Request Parameters
| Parameter | Type | Default | Description |
|---|---|---|---|
| model | string | — | Model ID (see Available Models above)*required |
| messages | array | — | Array of {role, content}. Roles: system, user, assistant*required |
| temperature | float | 0.7 | Sampling temperature 0–2. Lower = more focused |
| max_tokens | int | 4096 | Max tokens to generate |
| top_p | float | 1.0 | Nucleus sampling threshold |
| stream | bool | false | Stream response token-by-token |
| stop | string/array | null | Stop sequence(s) to end generation |
API Endpoints
/v1/chat/completionsSend a chat completion requestChat/v1/embeddingsGenerate text embeddingsEmbeddings/v1/modelsList all available modelsModels/auth/registerCreate a new accountAuth/auth/loginGet access tokenAuth/user/balanceCheck token balanceUser/user/subscriptionGet subscription detailsUser/api-keys/List your API keysAPI Keys/api-keys/Create a new API keyAPI Keys/api-keys/{key_id}Delete an API keyAPI KeysError Codes
Start building in minutes
100,000 free tokens. No credit card required.