awesome-free-llm-apis

Q: How do I install awesome-free-llm-apis?

Run npx skills add https://github.com/reason-machines/trending-skills --skill awesome-free-llm-apis in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/awesome-free-llm-apis/ (personal) or .claude/skills/awesome-free-llm-apis/ (project).

by reason-machinesAI & Agents

Reference guide for permanent free-tier LLM APIs with rate limits, model lists, and OpenAI-compatible integration patterns.

Installs: 982
Last 24h: —
Trending rank: —
Security: Not audited

Install

Works with every agent the skills CLI supports.

npx skills add https://github.com/reason-machines/trending-skills --skill awesome-free-llm-apis

01
Install
Run the command in your project. The CLI asks which agents to install for.
02
Describe the task
Ask your agent for something this skill covers. It loads awesome-free-llm-apis automatically when relevant.
03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.

SKILL.md

The instructions your agent reads when the skill is triggered.

Awesome Free LLM APIs

Skill by ara.so — Daily 2026 Skills collection.

A curated list of LLM providers offering permanent free tiers for text inference — no trial credits, no expiry. All endpoints listed are OpenAI SDK-compatible unless noted.

Provider Overview

Provider APIs (trained/fine-tuned by the company)

Provider	Notable Models	Rate Limits	Region
Cohere	Command A, Command R+, Aya Expanse 32B	20 RPM, 1K req/mo	🇺🇸
Google Gemini	Gemini 2.5 Pro, Flash, Flash-Lite	5–15 RPM, 100–1K RPD	🇺🇸 (not EU/UK/CH)
Mistral AI	Mistral Large 3, Small 3.1, Ministral 8B	1 req/s, 1B tok/mo	🇪🇺
Zhipu AI	GLM-4.7-Flash, GLM-4.5-Flash, GLM-4.6V-Flash	Undocumented	🇨🇳

Inference Providers (host open-weight models)

Provider	Notable Models	Rate Limits	Region
Cerebras	Llama 3.3 70B, Qwen3 235B, GPT-OSS-120B	30 RPM, 14,400 RPD	🇺🇸
Cloudflare Workers AI	Llama 3.3 70B, Qwen QwQ 32B	10K neurons/day	🇺🇸
GitHub Models	GPT-4o, Llama 3.3 70B, DeepSeek-R1	10–15 RPM, 50–150 RPD	🇺🇸
Groq	Llama 3.3 70B, Llama 4 Scout, Kimi K2	30 RPM, 1K RPD	🇺🇸
Hugging Face	Llama 3.3 70B, Qwen2.5 72B, Mistral 7B	$0.10/mo free credits	🇺🇸
Kluster AI	DeepSeek-R1, Llama 4 Maverick, Qwen3-235B	Undocumented	🇺🇸
LLM7.io	DeepSeek R1, Flash-Lite, Qwen2.5 Coder	30 RPM (120 with token)	🇬🇧
NVIDIA NIM	Llama 3.3 70B, Mistral Large, Qwen3 235B	40 RPM	🇺🇸
Ollama Cloud	DeepSeek-V3.2, Qwen3.5, Kimi-K2.5	1 concurrent, light usage	🇺🇸
OpenRouter	DeepSeek R1, Llama 3.3 70B, GPT-OSS-120B	20 RPM, 50 RPD (1K with $10+)	🇺🇸

Getting API Keys

Each provider has its own key management page:

# Store keys as environment variables — never hardcode them
export GROQ_API_KEY="your_groq_key"
export GEMINI_API_KEY="your_gemini_key"
export OPENROUTER_API_KEY="your_openrouter_key"
export MISTRAL_API_KEY="your_mistral_key"
export COHERE_API_KEY="your_cohere_key"
export CEREBRAS_API_KEY="your_cerebras_key"
export GITHUB_TOKEN="your_github_pat"
export HF_TOKEN="your_huggingface_token"
export NVIDIA_API_KEY="your_nvidia_key"
export CLOUDFLARE_API_TOKEN="your_cf_token"
export CLOUDFLARE_ACCOUNT_ID="your_cf_account_id"

OpenAI SDK Integration

All providers (except Ollama Cloud) are OpenAI SDK-compatible — just swap the base_url and api_key.

Python

from openai import OpenAI

# ── Groq ──────────────────────────────────────────────────────────────────────
client = OpenAI(
    base_url="https://api.groq.com/openai/v1",
    api_key=os.environ["GROQ_API_KEY"],
)
response = client.chat.completions.create(
    model="llama-3.3-70b-versatile",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

# ── Google Gemini ─────────────────────────────────────────────────────────────
client = OpenAI(
    base_url="https://generativelanguage.googleapis.com/v1beta/openai/",
    api_key=os.environ["GEMINI_API_KEY"],
)
response = client.chat.completions.create(
    model="gemini-2.0-flash",
    messages=[{"role": "user", "content": "Explain quantum entanglement."}],
)

# ── Mistral AI ────────────────────────────────────────────────────────────────
client = OpenAI(
    base_url="https://api.mistral.ai/v1",
    api_key=os.environ["MISTRAL_API_KEY"],
)
response = client.chat.completions.create(
    model="mistral-small-latest",
    messages=[{"role": "user", "content": "Write a haiku about code."}],
)

# ── OpenRouter ────────────────────────────────────────────────────────────────
client = OpenAI(
    base_url="https://openrouter.ai/api/v1",
    api_key=os.environ["OPENROUTER_API_KEY"],
)
response = client.chat.completions.create(
    model="deepseek/deepseek-r1",          # free model on OpenRouter
    messages=[{"role": "user", "content": "What is 2+2?"}],
    extra_headers={
        "HTTP-Referer": "https://yourapp.com",   # optional but recommended
        "X-Title": "My App",
    },
)

# ── Cerebras ──────────────────────────────────────────────────────────────────
client = OpenAI(
    base_url="https://api.cerebras.ai/v1",
    api_key=os.environ["CEREBRAS_API_KEY"],
)
response = client.chat.completions.create(
    model="llama-3.3-70b",
    messages=[{"role": "user", "content": "Tell me a joke."}],
)

# ── NVIDIA NIM ────────────────────────────────────────────────────────────────
client = OpenAI(
    base_url="https://integrate.api.nvidia.com/v1",
    api_key=os.environ["NVIDIA_API_KEY"],
)
response = client.chat.completions.create(
    model="meta/llama-3.3-70b-instruct",
    messages=[{"role": "user", "content": "Summarize this text."}],
)

# ── GitHub Models ─────────────────────────────────────────────────────────────
client = OpenAI(
    base_url="https://models.inference.ai.azure.com",
    api_key=os.environ["GITHUB_TOKEN"],
)
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Draft an email."}],
)

# ── Cohere (OpenAI-compatible endpoint) ───────────────────────────────────────
client = OpenAI(
    base_url="https://api.cohere.com/compatibility/v1",
    api_key=os.environ["COHERE_API_KEY"],
)
response = client.chat.completions.create(
    model="command-a-03-2025",
    messages=[{"role": "user", "content": "Translate to French: Hello world"}],
)

JavaScript / TypeScript

import OpenAI from "openai";

// ── Groq ──────────────────────────────────────────────────────────────────────
const groq = new OpenAI({
  baseURL: "https://api.groq.com/openai/v1",
  apiKey: process.env.GROQ_API_KEY,
});

const completion = await groq.chat.completions.create({
  model: "llama-3.3-70b-versatile",
  messages: [{ role: "user", content: "Hello!" }],
});
console.log(completion.choices[0].message.content);

// ── OpenRouter with free model router ────────────────────────────────────────
const openrouter = new OpenAI({
  baseURL: "https://openrouter.ai/api/v1",
  apiKey: process.env.OPENROUTER_API_KEY,
  defaultHeaders: {
    "HTTP-Referer": "https://yourapp.com",
    "X-Title": "My App",
  },
});

// Use the free models router — automatically picks an available free model
const freeCompletion = await openrouter.chat.completions.create({
  model: "openrouter/free",
  messages: [{ role: "user", content: "What is the capital of France?" }],
});

// ── Mistral ───────────────────────────────────────────────────────────────────
const mistral = new OpenAI({
  baseURL: "https://api.mistral.ai/v1",
  apiKey: process.env.MISTRAL_API_KEY,
});

const mistralCompletion = await mistral.chat.completions.create({
  model: "mistral-small-latest",
  messages: [{ role: "user", content: "Explain async/await in JavaScript." }],
});

Cloudflare Workers AI

Cloudflare uses a slightly different auth pattern:

import requests, os

ACCOUNT_ID = os.environ["CLOUDFLARE_ACCOUNT_ID"]
API_TOKEN  = os.environ["CLOUDFLARE_API_TOKEN"]

response = requests.post(
    f"https://api.cloudflare.com/client/v4/accounts/{ACCOUNT_ID}/ai/run/"
    "@cf/meta/llama-3.3-70b-instruct-fp8-fast",
    headers={"Authorization": f"Bearer {API_TOKEN}"},
    json={"messages": [{"role": "user", "content": "What is Cloudflare Workers?"}]},
)
result = response.json()
print(result["result"]["response"])

// Cloudflare Workers runtime (inside a Worker)
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const ai = new Ai(env.AI);
    const response = await ai.run("@cf/meta/llama-3.3-70b-instruct-fp8-fast", {
      messages: [{ role: "user", content: "Hello from Workers AI!" }],
    });
    return Response.json(response);
  },
};

Ollama Cloud (Non-OpenAI API)

Ollama Cloud uses the Ollama API format, not the OpenAI format:

import requests, os

response = requests.post(
    "https://ollama.com/api/chat",
    headers={"Authorization": f"Bearer {os.environ['OLLAMA_API_KEY']}"},
    json={
        "model": "deepseek-v3.2",
        "messages": [{"role": "user", "content": "What is 2 + 2?"}],
        "stream": False,
    },
)
print(response.json()["message"]["content"])

# Using the ollama Python client
import ollama, os

client = ollama.Client(
    host="https://ollama.com",
    headers={"Authorization": f"Bearer {os.environ['OLLAMA_API_KEY']}"},
)
response = client.chat(
    model="qwen3.5",
    messages=[{"role": "user", "content": "Write a poem about the sea."}],
)
print(response["message"]["content"])

Hugging Face Inference API

from openai import OpenAI
import os

client = OpenAI(
    base_url="https://router.huggingface.co/novita/v3/openai",
    api_key=os.environ["HF_TOKEN"],
)

response = client.chat.completions.create(
    model="meta-llama/llama-3.3-70b-instruct",
    messages=[{"role": "user", "content": "Summarize the theory of relativity."}],
    max_tokens=512,
)
print(response.choices[0].message.content)

Streaming Responses

from openai import OpenAI
import os

client = OpenAI(
    base_url="https://api.groq.com/openai/v1",
    api_key=os.environ["GROQ_API_KEY"],
)

with client.chat.completions.stream(
    model="llama-3.3-70b-versatile",
    messages=[{"role": "user", "content": "Write a short story about a robot."}],
) as stream:
    for text in stream.text_stream:
        print(text, end="", flush=True)

const stream = await groq.chat.completions.create({
  model: "llama-3.3-70b-versatile",
  messages: [{ role: "user", content: "Write a haiku." }],
  stream: true,
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content ?? "");
}

Provider Fallback Pattern

Cycle through providers when rate limits are hit:

from openai import OpenAI, RateLimitError
import os

PROVIDERS = [
    {
        "name": "Groq",
        "base_url": "https://api.groq.com/openai/v1",
        "api_key": os.environ.get("GROQ_API_KEY"),
        "model": "llama-3.3-70b-versatile",
    },
    {
        "name": "Cerebras",
        "base_url": "https://api.cerebras.ai/v1",
        "api_key": os.environ.get("CEREBRAS_API_KEY"),
        "model": "llama-3.3-70b",
    },
    {
        "name": "Mistral",
        "base_url": "https://api.mistral.ai/v1",
        "api_key": os.environ.get("MISTRAL_API_KEY"),
        "model": "mistral-small-latest",
    },
    {
        "name": "OpenRouter",
        "base_url": "https://openrouter.ai/api/v1",
        "api_key": os.environ.get("OPENROUTER_API_KEY"),
        "model": "openrouter/free",
    },
]

def chat_with_fallback(messages: list[dict], **kwargs) -> str:
    for provider in PROVIDERS:
        if not provider["api_key"]:
            continue
        try:
            client = OpenAI(
                base_url=provider["base_url"],
                api_key=provider["api_key"],
            )
            response = client.chat.completions.create(
                model=provider["model"],
                messages=messages,
                **kwargs,
            )
            return response.choices[0].message.content
        except RateLimitError:
            print(f"Rate limited on {provider['name']}, trying next...")
            continue
        except Exception as e:
            print(f"Error on {provider['name']}: {e}, trying next...")
            continue
    raise RuntimeError("All providers exhausted.")

# Usage
answer = chat_with_fallback(
    messages=[{"role": "user", "content": "What is the speed of light?"}]
)
print(answer)

OpenRouter Free Models Router

OpenRouter provides a special router that automatically selects available free models:

from openai import OpenAI
import os

client = OpenAI(
    base_url="https://openrouter.ai/api/v1",
    api_key=os.environ["OPENROUTER_API_KEY"],
)

# Use the free router — picks from 29+ free models automatically
response = client.chat.completions.create(
    model="openrouter/free",
    messages=[{"role": "user", "content": "Explain recursion."}],
)

# Or use model fallbacks for priority ordering
response = client.chat.completions.create(
    model="deepseek/deepseek-r1",
    messages=[{"role": "user", "content": "Explain recursion."}],
    extra_body={
        "route": "fallback",
        "models": [
            "deepseek/deepseek-r1",
            "meta-llama/llama-3.3-70b-instruct:free",
            "openrouter/free",
        ],
    },
)

LangChain Integration

from langchain_openai import ChatOpenAI
from langchain_core.messages import HumanMessage
import os

# Works with any OpenAI-compatible provider
llm = ChatOpenAI(
    model="llama-3.3-70b-versatile",
    openai_api_base="https://api.groq.com/openai/v1",
    openai_api_key=os.environ["GROQ_API_KEY"],
    temperature=0.7,
)

response = llm.invoke([HumanMessage(content="What are the SOLID principles?")])
print(response.content)

# Gemini via LangChain
gemini = ChatOpenAI(
    model="gemini-2.0-flash",
    openai_api_base="https://generativelanguage.googleapis.com/v1beta/openai/",
    openai_api_key=os.environ["GEMINI_API_KEY"],
)

Rate Limit Reference

Provider	RPM	RPD	Notes
Groq	30	1,000	14,400 RPD for Llama 3.1 8B only
Cerebras	30	14,400	—
Gemini Flash	15	1,500	Not in EU/UK/CH
Gemini 2.5 Pro	5	25	Not in EU/UK/CH
GitHub Models	10–15	50–150	Varies by model tier
OpenRouter (free)	20	50	1K RPD after $10+ purchase
Mistral	1 req/s	—	1B tokens/month cap
NVIDIA NIM	40	—	—
Cloudflare Workers AI	—	—	10K neurons/day
Cohere	20	—	1K requests/month

Common Troubleshooting

AuthenticationError

Double-check the env var is set: echo $GROQ_API_KEY
Ensure the key is for the correct provider
Some providers (GitHub Models) require a classic PAT, not a fine-grained token

RateLimitError

Implement exponential backoff or use the fallback pattern above
Switch to a provider with higher limits (Cerebras: 14,400 RPD)
For Groq, use llama-3.1-8b-instant for the 14,400 RPD limit

Model not found

Check the exact model ID on the provider's docs/dashboard
OpenRouter free models have :free suffix: meta-llama/llama-3.3-70b-instruct:free
Cloudflare models use @cf/ prefix: @cf/meta/llama-3.3-70b-instruct-fp8-fast

Gemini free tier unavailable

The free tier is not available in EU, UK, or Switzerland
Use a VPN or switch to a different provider like Groq or Mistral

Ollama Cloud not working with OpenAI SDK

Ollama Cloud uses its own API format — use the ollama Python package or raw HTTP

OpenRouter 50 RPD limit

Make a one-time $10 credit purchase to unlock 1,000 RPD for free models permanently
Alternatively, use openrouter/free router to distribute across all free models

Choosing the Right Provider

Need highest RPD?         → Cerebras (14,400 RPD)
Need smartest free model? → Gemini 2.5 Pro (if not in EU/UK/CH)
Need EU-hosted?           → Mistral AI (France)
Need most model variety?  → OpenRouter (29+ free models) or Cloudflare (48+ models)
Need fastest inference?   → Groq (purpose-built inference chips)
Need reasoning model?     → DeepSeek-R1 on Groq/OpenRouter/Kluster AI
Need vision?              → Gemini Flash, Llama 4 Scout (Groq), GLM-4.6V-Flash (Zhipu)
No rate limit concern?    → Cloudflare (10K neurons/day, compute-based)

FAQ

What does the awesome-free-llm-apis skill do?

Reference guide for permanent free-tier LLM APIs with rate limits, model lists, and OpenAI-compatible integration patterns.

How do I install awesome-free-llm-apis?

Run npx skills add https://github.com/reason-machines/trending-skills --skill awesome-free-llm-apis in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/awesome-free-llm-apis/ (personal) or .claude/skills/awesome-free-llm-apis/ (project).

Which AI agents can use awesome-free-llm-apis?

Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.

Is awesome-free-llm-apis safe to install?

No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.

Who maintains awesome-free-llm-apis?

It is published by reason-machines in reason-machines/trending-skills. It has 982 recorded installs.

Popular ai & agents skills

View category →

google-agents-cli-publish

google/agents-cli

This skill should be used when the user wants to "publish an agent", "publish my ADK agent", "register an agent with Gemini", "publish to Gemini" (formerly Gemini Enterprise), or needs guidance on the agents-cli publish gemini command. Also use when the user wants to "manage agents in Agent Registry", "list/update/delete registered agents", or "register an MCP server". Covers ADK vs A2A registration modes, programmatic and interactive usage, flag reference, auto-detection from deployment metadata, Agent Registry fleet management, and troubleshooting. Part of the agents-cli skills suite. Do NOT use for deployment (use google-agents-cli-deploy).

AI & Agents513.9K installs

google-agents-cli-eval

google/agents-cli

This skill should be used when the user wants to "run an evaluation", "evaluate my agent", "evaluate my ADK agent", "write an eval dataset", "analyze eval failures", "compare eval results", "optimize agent", or needs guidance on the Gemini platform eval methodology and the Quality Flywheel. Covers eval metrics, dataset schema, LLM-as-judge scoring, and common failure causes. Applies to any agents-cli project, whatever framework the agent is written in. Do NOT use for agent API code patterns (ADK: use google-agents-cli-adk-code), deployment (use google-agents-cli-deploy), or project scaffolding (use google-agents-cli-scaffold).

AI & Agents513.5K installs

skill-creator

anthropics/skills

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.

OfficialAI & Agents416.4K installs

codex-pet

prime-skills/runcomfy-agent-skills

Codex Pet generator on RunComfy. Build a Codex-compatible Codex Pet spritesheet.webp + pet.json from a single reference image, drop it into ${CODEXHOME:-$HOME/.codex}/pets/ / and Codex picks it up as a custom Codex Pet next to the 8 built-ins. This skill produces the exact Codex Pet atlas Codex expects (1536x1872 PNG/WebP, 8 cols x 9 rows, 192x208 cells, 9 animation states — idle, running-right, running-left, waving, jumping, failed, waiting, running, review). Calls OpenAI GPT Image 2 edit ONCE via the local RunComfy CLI as runcomfy run openai/gpt-image-2/edit to produce a canonical Codex Pet pose, then assembles all 9 animation rows programmatically with ImageMagick micro-transforms — no Codex Pro, no $imagegen, no OPENAIAPIKEY required, only RUNCOMFYTOKEN. Triggers on "codex pet", "create codex pet", "make codex pet", "hatch codex pet", "/hatch image", "desktop pet codex", "codex pets", "spritesheet.webp", or any explicit ask to build a custom pet for OpenAI Codex.

AI & Agents402.1K installs

full-output-enforcement

leonxlnx/taste-skill

Overrides default LLM truncation behavior. Enforces complete code generation, bans placeholder patterns, and handles token-limit splits cleanly. Apply to any task requiring exhaustive, unabridged output.

AI & Agents401K installs

executing-plans

obra/superpowers

Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available

AI & Agents245.7K installs

awesome-free-llm-apis

Install

Install

Describe the task

Review the result

SKILL.md

Awesome Free LLM APIs

Provider Overview

Provider APIs (trained/fine-tuned by the company)

Inference Providers (host open-weight models)

Getting API Keys

OpenAI SDK Integration

Python

JavaScript / TypeScript

Cloudflare Workers AI

Ollama Cloud (Non-OpenAI API)

Hugging Face Inference API

Streaming Responses

Provider Fallback Pattern

OpenRouter Free Models Router

LangChain Integration

Rate Limit Reference

Common Troubleshooting

Choosing the Right Provider

FAQ

More from reason-machines/trending-skills

openclaw-control-center

inkos-multi-agent-novel-writing

claude-hud-statusline

understand-anything-knowledge-graph

agency-agents-ai-specialists

agent-browser-automation

Popular ai & agents skills

google-agents-cli-publish

google-agents-cli-eval

skill-creator

codex-pet

full-output-enforcement

executing-plans