Hugging Face

Hugging Face Inference Providers ist ein Router vor vielen Hosts. Ein HF_TOKEN, eine router.huggingface.co/v1-URL und $0,10 Guthaben pro Monat auf dem Gratis-Konto. Der Live-Router listet 135 Modell-ids — keine feste „Gratis-Seven“.

API-Schlüssel holen Offizielle Doku ↗

Der Router wählt den Anbieter. Ein :provider an der Modell-id nagt einen fest, damit nicht der Überraschungsgast die Anfrage ausführt. Der OpenAI-kompatible Pfad deckt nur Chat ab; Embeddings und der Rest laufen über den Hugging-Face-Client.

Kostenlose Modelle: 135
Max. Kontext: 131K
Gratis-Tarif: $0,10 Guthaben/Monat
Anforderung: E-Mail-Konto, keine Karte

Endpunkt

Basis-URLhttps://router.huggingface.co/v1

UmgebungsvariableHF_TOKEN

Ratengrenze$0.10 of credits per month

OpenAI-kompatibel
Eigenes SDK: Hugging Face Inference
Text
Code

Codebeispiele

npm install openai

import OpenAI from 'openai'

const client = new OpenAI({
  baseURL: 'https://router.huggingface.co/v1',
  apiKey: process.env.HF_TOKEN
})

const response = await client.chat.completions.create({
  model: 'zai-org/GLM-5.3-Flash',
  messages: [{ role: 'user', content: 'Explain closures in one paragraph.' }]
})

console.log(response.choices[0].message.content)

pip install openai

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://router.huggingface.co/v1",
    api_key=os.environ["HF_TOKEN"],
)

response = client.chat.completions.create(
    model="zai-org/GLM-5.3-Flash",
    messages=[{"role": "user", "content": "Explain closures in one paragraph."}],
)

print(response.choices[0].message.content)

curl https://router.huggingface.co/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $HF_TOKEN" \
  -d '{
    "model": "zai-org/GLM-5.3-Flash",
    "messages": [{"role": "user", "content": "Explain closures in one paragraph."}]
  }'

npm install @huggingface/inference

import { InferenceClient } from '@huggingface/inference'

const client = new InferenceClient(process.env.HF_TOKEN)

const response = await client.chatCompletion({
  model: 'zai-org/GLM-5.3-Flash',
  messages: [{ role: 'user', content: 'Explain closures in one paragraph.' }]
})

console.log(response.choices[0].message.content)

pip install huggingface_hub

import os
from huggingface_hub import InferenceClient

client = InferenceClient(token=os.environ["HF_TOKEN"])

response = client.chat_completion(
    model="zai-org/GLM-5.3-Flash",
    messages=[{"role": "user", "content": "Explain closures in one paragraph."}],
)

print(response.choices[0].message.content)

Wissenswert

Der Router wählt den Anbieter; mit :provider an der Modell-id wird einer festgenagelt.
Der Gratis-Anteil ist ein monatliches Guthaben von $0,10 über den ganzen Router, keine feste Liste von 7 Modellen.

Modelle, die Sie gratis bekommen135

GLM 5.3 Flashzai-org/GLM-5.3-Flash
GLM 5.3zai-org/GLM-5.3
Qwen 3.8 27BQwen/Qwen3.8-27B
Kimi K3moonshotai/Kimi-K3
DeepSeek V4 Flash 0731deepseek-ai/DeepSeek-V4-Flash-0731
Granite 4.2 30Bibm-granite/granite-4.2-30b
GLM 5.3 Flash BF16zai-org/GLM-5.3-Flash-BF16
Muse Glimmer 30Bmeta-models/Muse-Glimmer-30B
Gemma 4 31Bgoogle/gemma-4-31B-it
Granite 4.2 3Bibm-granite/granite-4.2-3b
Llama 3.1 8Bmeta-llama/Llama-3.1-8B-Instruct
Granite 4.2 8Bibm-granite/granite-4.2-8b
GLM 5.3 BF16zai-org/GLM-5.3-BF16
Qwen 3.8 2.4T A95BQwen/Qwen3.8-2.4T-A95B
DeepSeek V4 Pro 0813deepseek-ai/DeepSeek-V4-Pro-0813
DeepSeek V4 Flashdeepseek-ai/DeepSeek-V4-Flash
Qwen 3.5 9BQwen/Qwen3.5-9B
Inkling Smallthinkingmachines/Inkling-Small
Qwen 3.6 35B A3BQwen/Qwen3.6-35B-A3B
Gemma 4 26B A4Bgoogle/gemma-4-26B-A4B-it
Hy3tencent/Hy3
GLM 5.2zai-org/GLM-5.2
Inklingthinkingmachines/Inkling
Ling 3.0 FlashinclusionAI/Ling-3.0-flash
GPT-OSS 20Bopenai/gpt-oss-20b
GPT-OSS 120Bopenai/gpt-oss-120b
MiniMax M3MiniMaxAI/MiniMax-M3
DeepSeek V4 Prodeepseek-ai/DeepSeek-V4-Pro
Ternary Bonsai 27B Ggufprism-ml/Ternary-Bonsai-27B-gguf
GLM 4.7 Flashzai-org/GLM-4.7-Flash
DeepSeek R1deepseek-ai/DeepSeek-R1
Qwen 3 8BQwen/Qwen3-8B
Llama 3.3 70Bmeta-llama/Llama-3.3-70B-Instruct
Qwen 2.5 Coder 7BQwen/Qwen2.5-Coder-7B-Instruct
NVIDIA Nemotron 3 Ultra 550B A55B NVFP4nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-NVFP4
Qwen 3 Coder NextQwen/Qwen3-Coder-Next
NVIDIA Nemotron 3.5 Lightning 30B A3B BF16nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16
Qwen 3 Coder 30B A3BQwen/Qwen3-Coder-30B-A3B-Instruct
Qwen 3 4B 2507Qwen/Qwen3-4B-Instruct-2507
Qwen 3.5 35B A3BQwen/Qwen3.5-35B-A3B
Qwen 3.6 27BQwen/Qwen3.6-27B
MiMo V2.5XiaomiMiMo/MiMo-V2.5
Qwen 2.5 Coder 32BQwen/Qwen2.5-Coder-32B-Instruct
Apertus V1.5 8Bswiss-ai/Apertus-v1.5-8B
Qwen 3 14BQwen/Qwen3-14B
Kimi K2.6moonshotai/Kimi-K2.6
GLM 5.2 FP8zai-org/GLM-5.2-FP8
Kimi K2.7 Codemoonshotai/Kimi-K2.7-Code
Phi 4microsoft/phi-4
Gemma 3 4Bgoogle/gemma-3-4b-it
Qwen 3 235B A22B 2507Qwen/Qwen3-235B-A22B-Instruct-2507
Apertus V1.5 70Bswiss-ai/Apertus-v1.5-70B
L3 8B Stheno V3.2Sao10K/L3-8B-Stheno-v3.2
Qwen 3 30B A3BQwen/Qwen3-30B-A3B
Step 3.7 Flashstepfun-ai/Step-3.7-Flash
Qwen 3 Next 80B A3BQwen/Qwen3-Next-80B-A3B-Instruct
DeepSeek V3.2deepseek-ai/DeepSeek-V3.2
MiMo V2.5 ProXiaomiMiMo/MiMo-V2.5-Pro
Llama 4 Scout 17B 16Emeta-llama/Llama-4-Scout-17B-16E-Instruct
Qwen 3 32BQwen/Qwen3-32B
Qwen 3.5 397B A17BQwen/Qwen3.5-397B-A17B
Qwen 3 235B A22BQwen/Qwen3-235B-A22B
Kimi K2.5moonshotai/Kimi-K2.5
Gemma 3 27Bgoogle/gemma-3-27b-it
Qwen 3.5 27BQwen/Qwen3.5-27B
Qwen 3.5 122B A10BQwen/Qwen3.5-122B-A10B
Qwen 3 235B A22B Thinking 2507Qwen/Qwen3-235B-A22B-Thinking-2507
Tiny Aya GlobalCohereLabs/tiny-aya-global
GLM 5zai-org/GLM-5
Gemma 3 12Bgoogle/gemma-3-12b-it
Hermes 3 Llama 3.1 70BNousResearch/Hermes-3-Llama-3.1-70B
Tiny Aya WaterCohereLabs/tiny-aya-water
GLM 5.1zai-org/GLM-5.1
Llama Guard 4 12Bmeta-llama/Llama-Guard-4-12B
Qwen 2.5 72BQwen/Qwen2.5-72B-Instruct
GLM 5.1 FP8zai-org/GLM-5.1-FP8
Apertus 8B 2509swiss-ai/Apertus-8B-Instruct-2509
DeepSeek V3deepseek-ai/DeepSeek-V3
GLM 4.7zai-org/GLM-4.7
DeepSeek V3 0324deepseek-ai/DeepSeek-V3-0324
DeepSeek R1 Distill Llama 8Bdeepseek-ai/DeepSeek-R1-Distill-Llama-8B
WizardLM 2 8x22Balpindale/WizardLM-2-8x22B
ERNIE 4.5 VL 424B A47B Base PTbaidu/ERNIE-4.5-VL-424B-A47B-Base-PT
DeepSeek V3.2 Expdeepseek-ai/DeepSeek-V3.2-Exp
GLM 4.6V Flashzai-org/GLM-4.6V-Flash
AutoGLM Phone 9B Multilingualzai-org/AutoGLM-Phone-9B-Multilingual
Ternary Bonsai 27B AWQ 4bitprism-ml/Ternary-Bonsai-27B-AWQ-4bit
L3 8B Lunaris V1Sao10K/L3-8B-Lunaris-v1
Gemma SEA LION V4 27Baisingapore/Gemma-SEA-LION-v4-27B-IT
Llama 4 Maverick 17B 128E FP8meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
Qwen 3 VL 235B A22BQwen/Qwen3-VL-235B-A22B-Instruct
DeepSeek R1 Distill Llama 70Bdeepseek-ai/DeepSeek-R1-Distill-Llama-70B
Apertus 70B 2509swiss-ai/Apertus-70B-Instruct-2509
GLM 4.5Vzai-org/GLM-4.5V
Qwen 3 VL 30B A3BQwen/Qwen3-VL-30B-A3B-Instruct
GLM 4 32B 0414zai-org/GLM-4-32B-0414
Qwen SEA LION V4 32Baisingapore/Qwen-SEA-LION-v4-32B-IT
Qwen 3 VL 235B A22B ThinkingQwen/Qwen3-VL-235B-A22B-Thinking
Qwen 2.5 Coder 3BQwen/Qwen2.5-Coder-3B-Instruct
MiniMax M1 80kMiniMaxAI/MiniMax-M1-80k
Bielik 11B V3.0speakleash/Bielik-11B-v3.0-Instruct
Qwen 2.5 VL 72BQwen/Qwen2.5-VL-72B-Instruct
DeepSeek R1 Distill Qwen 7Bdeepseek-ai/DeepSeek-R1-Distill-Qwen-7B
Kimi K2moonshotai/Kimi-K2-Instruct
DeepSeek R1 0528deepseek-ai/DeepSeek-R1-0528
GLM 4.5zai-org/GLM-4.5
DeepSeek R1 Distill Qwen 14Bdeepseek-ai/DeepSeek-R1-Distill-Qwen-14B
GLM 4.5 Airzai-org/GLM-4.5-Air
DeepSeek V3.1deepseek-ai/DeepSeek-V3.1
GLM 4.6 FP8zai-org/GLM-4.6-FP8
Qwen 3 4B Thinking 2507Qwen/Qwen3-4B-Thinking-2507
Qwen 3 Coder 480B A35BQwen/Qwen3-Coder-480B-A35B-Instruct
DeepSeek V3.1 Terminusdeepseek-ai/DeepSeek-V3.1-Terminus
C4ai Command R 08 2024CohereLabs/c4ai-command-r-08-2024
GLM 4.6zai-org/GLM-4.6
Kimi K2 0905moonshotai/Kimi-K2-Instruct-0905
Step 3.5 Flashstepfun-ai/Step-3.5-Flash
Aya Expanse 32BCohereLabs/aya-expanse-32b
GLM 4.7 FP8zai-org/GLM-4.7-FP8
MiniMax M2MiniMaxAI/MiniMax-M2
MiniMax M2.7MiniMaxAI/MiniMax-M2.7
C4ai Command R7B 12 2024CohereLabs/c4ai-command-r7b-12-2024
MiniMax M2.1MiniMaxAI/MiniMax-M2.1
GLM 4.5V FP8zai-org/GLM-4.5V-FP8
C4ai Command R7B Arabic 02 2025CohereLabs/c4ai-command-r7b-arabic-02-2025
MiniMax M2.5MiniMaxAI/MiniMax-M2.5
GLM 4.6Vzai-org/GLM-4.6V
C4ai Command A 03 2025CohereLabs/c4ai-command-a-03-2025
Aya Vision 32BCohereLabs/aya-vision-32b
GLM 4.6V FP8zai-org/GLM-4.6V-FP8
GPT-OSS Safeguard 20Bopenai/gpt-oss-safeguard-20b
Command A Reasoning 08 2025CohereLabs/command-a-reasoning-08-2025
Command A Translate 08 2025CohereLabs/command-a-translate-08-2025
Tiny Aya EarthCohereLabs/tiny-aya-earth
Tiny Aya FireCohereLabs/tiny-aya-fire