Updated daily
Every AI API.
Live-tested. Free tiers first.
The only AI API directory that checks if APIs are actually working today and shows you how to use them in one click.
12
APIs tracked
12
Live right now
9
Free tier
24h
Check interval
What are you building?Pick a use case for a ranked shortlist
⌕⌘K
All APIs
Sorted by free tier, then popularity
Groq
inference
Live
● Live✦ Free tier~206ms$0.05–$0.59 / 1M tokens
Ultra-fast LLM inference using LPU hardware. Supports Llama 3, Mixtral, and Gemma.
✦14,400 req/day, 30 req/min
checked 2h ago
llama3mixtralfast
206ms avg
from groq import Groq
client = Groq(api_key="YOUR_KEY")
res = client.chat.completions.create(
model="llama3-8b-8192",
messages=[{"role": "user", "content": "Hello"}]
)
print(res.choices[0].message.content)Mistral
inference
Live
● Live✦ Free tier~315ms$0.20–$2.00 / 1M tokens
European open-weight LLMs with no daily request cap on the free tier.
✦1 req/sec, no daily cap
checked 2h ago
mistral-7bopen-weight
315ms avg
from mistralai import Mistral
client = Mistral(api_key="YOUR_KEY")
res = client.chat.complete(
model="mistral-small-latest",
messages=[{"role": "user", "content": "Hello"}]
)
print(res.choices[0].message.content)Hugging Face
inference
Live
● Live✦ Free tier~49msPay-as-you-go, varies by model
Access thousands of open-source models via a single OpenAI-compatible API. Free with a Hugging Face token.
✦Free with HF token, rate-limited
checked 2h ago
multi-modelopen-source
49ms avg
from huggingface_hub import InferenceClient
client = InferenceClient(api_key="YOUR_HF_TOKEN")
res = client.chat.completions.create(
model="meta-llama/Meta-Llama-3-8B-Instruct",
messages=[{"role": "user", "content": "Hello"}],
max_tokens=500
)
print(res.choices[0].message.content)Together AI
inference
Live
● Live✦ Free tier~136ms$0.10–$0.90 / 1M tokens
Run 50+ open-source models. Competitive pay-as-you-go pricing with a signup credit.
✦$1 credit on signup
checked 2h ago
llama3qwenmulti-model
136ms avg
from together import Together
client = Together(api_key="YOUR_KEY")
res = client.chat.completions.create(
model="meta-llama/Llama-3-8b-chat-hf",
messages=[{"role": "user", "content": "Hello"}]
)
print(res.choices[0].message.content)Anthropic
inference
Live
● Live✕ No free tier~33ms$3 / 1M input, $15 / 1M output (Sonnet)
Claude model family. Best-in-class reasoning and 200k token context window.
✕$3/M input tokens (Sonnet)
checked 2h ago
claudereasoninglong-ctx
33ms avg
import anthropic
client = anthropic.Anthropic(api_key="YOUR_KEY")
msg = client.messages.create(
model="claude-sonnet-4-6",
max_tokens=1024,
messages=[{"role": "user", "content": "Hello"}]
)
print(msg.content[0].text)OpenAI
inference
Live
● Live✕ No free tier~37ms$0.15 / 1M input, $0.60 / 1M output (4o-mini)
GPT-4o and o-series models. Industry standard with the widest ecosystem support.
✕$0.005/1K tokens (gpt-4o-mini)
checked 2h ago
gpt-4oo3vision
37ms avg
from openai import OpenAI
client = OpenAI(api_key="YOUR_KEY")
res = client.chat.completions.create(
model="gpt-4o-mini",
messages=[{"role": "user", "content": "Hello"}]
)
print(res.choices[0].message.content)Cohere
embeddings
Live
● Live✕ No free tier~271ms$0.10 / 1M tokens (embed v3)
Enterprise RAG and embeddings. Best-in-class reranking and multilingual support.
✕Free tier ended Jan 2025
checked 2h ago
ragembedrerank
271ms avg
import cohere
co = cohere.Client("YOUR_KEY")
res = co.embed(
texts=["Hello world"],
model="embed-english-v3.0",
input_type="search_document"
)
print(res.embeddings[0][:5])Voyage AI
embeddings
Live
● Live✦ Free tier~72ms$0.06–$0.18 / 1M tokens
State-of-the-art embedding models. Used by Anthropic for RAG use cases.
✦50M tokens free per month
checked 2h ago
embeddingsragretrieval
72ms avg
import voyageai
vo = voyageai.Client(api_key="YOUR_KEY")
result = vo.embed(
["Hello world"],
model="voyage-3",
input_type="document"
)
print(result.embeddings[0][:5])Stability AI
image-gen
Live
● Live✦ Free tier~80ms~$0.01–$0.04 per image (credit-based)
Stable Diffusion and SDXL image generation with high quality customizable outputs.
✦25 free credits on signup
checked 2h ago
stable-diffusionsdxlimage
80ms avg
import requests, base64
res = requests.post(
"https://api.stability.ai/v1/generation/stable-diffusion-v1-6/text-to-image",
headers={"Authorization": "Bearer YOUR_KEY"},
json={"text_prompts": [{"text": "a cat in space"}], "samples": 1}
)
with open("out.png", "wb") as f:
f.write(base64.b64decode(res.json()["artifacts"][0]["base64"]))Replicate
image-gen
Live
● Live✦ Free tier~351msPer-second compute, ~$0.003 per image (Flux Schnell)
Run any open-source AI model via API. Covers image, video, audio, and language models.
✦$5 free credit on signup
checked 2h ago
sdxlfluxmulti-model
351ms avg
import replicate
output = replicate.run(
"black-forest-labs/flux-schnell",
input={"prompt": "a cat in space"}
)
print(output)Google Gemini
multimodal
Live
● Live✦ Free tier~115ms$0.075 / 1M input, $0.30 / 1M output (1.5 Flash)
Gemini 1.5 Pro and Flash models. 1M token context, vision, audio, and code support.
✦15 RPM, 1M TPM on Flash
checked 2h ago
geminivisionlong-ctx
115ms avg
import google.generativeai as genai
genai.configure(api_key="YOUR_KEY")
model = genai.GenerativeModel("gemini-1.5-flash")
res = model.generate_content("Hello")
print(res.text)DeepSeek
inference
Live
● Live✦ Free tier~104ms$0.27 / 1M input, $1.10 / 1M output
High performance open-source models at very low cost. Strong on coding and math.
✦500M tokens free on signup
checked 2h ago
deepseek-v3codingcheap
104ms avg
from openai import OpenAI
client = OpenAI(api_key="YOUR_KEY", base_url="https://api.deepseek.com")
res = client.chat.completions.create(
model="deepseek-chat",
messages=[{"role": "user", "content": "Hello"}]
)
print(res.choices[0].message.content)