Skip to content

ShaikhWarsi/free-ai-tools

BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
website
website
 
 
CONTRIBUTING.md
CONTRIBUTING.md
 
 
LICENSE
LICENSE
 
 
README.md
README.md
 
 

Repository files navigation

Free AI Tools

Stars Last Updated License Contributions

Curated list of free LLM APIs, coding copilots, AI IDEs, agents, and infrastructure tools for building real AI applications.

What's Inside

  • ✅ Free GPT-5 / Claude / Gemini API access
  • 🤖 Coding copilots and AI-native IDEs (Cursor, Trae, Windsurf)
  • 💰 Cheapest AI APIs ($0.10-0.50 per 1M tokens)
  • 📚 RAG stack tools (vector DBs, embeddings, frameworks)
  • 🎯 Agent frameworks and automation tools
  • 🔒 Local models for privacy (Ollama, Llama, Qwen)
  • 🏗️ Production-ready stack configurations

Goal: Help developers build AI apps without paying $200/month.

Note

Please don't abuse these services, else we might lose them for everyone.

Warning

April 2026 Model Tier Changes: Major providers (OpenAI, Anthropic, Google) have restricted flagship models (GPT-5.4, Claude Opus 4.6, Gemini 3.1 Pro) to paid tiers. Free tiers now get lighter models (GPT-4o, Claude Sonnet/Haiku, Gemini Flash). Entries marked with [verify] need confirmation.

🎯 Why This Repo Exists

Most AI tool lists are:

  • ❌ Outdated (prices/limits from 2023)
  • ❌ Filled with affiliate links and sponsored placements
  • ❌ General-purpose directories with no developer focus
  • ❌ Missing production-critical details (rate limits, commercial use, architecture patterns)

This repo focuses only on:

  • ✅ Tools developers actually use in production
  • ✅ Generous free tiers (no "5 requests then paywall")
  • ✅ Production-capable models (SWE-bench verified, not toys)
  • ✅ Real infrastructure (APIs, hosting, vector DBs, not just chatbots)
  • ✅ Minimal fluff, maximum utility

Unlike: awesome-ai (general list), ai-collection (marketing focus), toolify (affiliate-heavy)

This is for: Builders who want to ship AI features this week.

⭐ Support This Project

If this repo helped you build something or saved you money:

⭐ Star this repo — it helps more builders discover free AI resources.

[🔄 Share with your team] — spread the knowledge.

📝 Contribute — found a new free tier? Updated pricing? PRs welcome!

📅 Updates

2026-04-12

  • ✨ added a website for easy navigation

2026-04-11

  • ✨ Initial release

Table of Contents

Quick Comparison

Free LLM API Providers Summary

Provider Models Free Tier Credit Card
NVIDIA NIM 46 40 req/min No
OpenRouter 25 50/day (1K/day with $10) No
Groq 20+ 1K-14.4K req/day No
Google AI Studio 9 5-500 req/day No
Cloudflare Workers AI 47+ 10K neurons/day No
Cerebras 4 1M tokens/day No
Cohere 14 1K req/month No
Mistral La Plateforme 10+ 1B tokens/month No
GitHub Models 30+ 50 chat + 2K completions/month No
SambaNova 13 $5 for 3 months No
Hyperbolic 13 $1 trial No

AI-Powered IDEs with Free Pro-Grade Access

IDE Pro-grade Models Free Tier Limit Credit Card
Cursor GPT-5.1-Codex-Max Limited free tier No
Trae DeepSeek V4, GPT-4.1 (Claude removed Nov 2025) 10 fast + 50 slow/month No
Windsurf OpenAI, Anthropic, Google, xAI 25 credits/month Required
Qoder Qwen3.6-Plus, Qwen3-Coder-480B, Claude, GPT, Gemini Unlimited completions + limited chat No

CLI Coding Tools with Free Pro-Grade Access

Tool Pro-grade Models Free Tier Limit Credit Card
Gemini CLI Gemini 3.1 Flash [verify: Pro paid] 100-250 req/day No
Rovo Dev CLI Claude Sonnet 4 [verify], GPT-5 preview [verify] 5M tokens/day No
Warp GPT-4.1, Claude Opus 4.1 [verify] 150 credits/month No
GitHub Copilot GPT-4.1, Claude Opus 50 chat + 2K completions/month No
Jules Gemini 2.5 Pro 15 tasks/day No
AWS Kiro Claude Sonnet 4 [verify] 50 credits/month No
OpenCode 300+ models via OpenRouter Zen Free tier No
ForgeCode 300+ models via OpenRouter 10K tokens/day No
Amazon Q Developer Claude Sonnet 4 [verify] 50 agentic req/month Required
RooCode Bring your own keys Unlimited (BYOK) No
Goose Bring your own keys Unlimited (BYOK) No
OhMyPi Bring your own keys Unlimited (BYOK) No

What Qualifies as "Pro-Grade"?

Models achieving ≥60% on SWE-bench Verified:

Model SWE-bench Provider
Claude Opus 4.6 84.2% Anthropic
GPT-5.4 80.1% OpenAI
Claude Sonnet 4.6 79.3% Anthropic
Gemini 3.1 Pro 77.4% Google
Claude Opus 4.5 82.1% Anthropic
GPT-5.1-Codex-Max 78.3% OpenAI
Qwen3.6-Plus 71.2% Alibaba
Claude Sonnet 4.5 77.8% Anthropic

Note: [verify] indicates scores need verification from official sources. Always check current benchmarks before making decisions.

🏗️ Recommended Stacks

Ready-made combinations for different use cases. Copy-paste these configurations.

🟢 Fully Free Coding Stack (No Credit Card)

Layer Tool Why
IDE Cursor Hobby / Qoder GPT-5.4 limited credits
CLI Gemini CLI (3.1 Pro) / Rovo 100-250 req/day, 5M tokens/day
API OpenRouter + Groq 50 req/day + 14.4K req/day combo
Local Ollama + Qwen3.6-Plus Unlimited offline
Automation n8n Self-hosted Unlimited workflows
Vector DB ChromaDB / LanceDB Free local storage

Total Cost: $0/month

⚡ Fastest Stack (Low Latency)

Layer Tool Speed
Inference Groq / Cerebras 2,000 tokens/sec (Cerebras)
Coding Qwen3.6-Plus via Groq 1,000 req/day (71.2% SWE)
Agent OpenCode Zen Big Pickle (72.0%), MiniMax M2.5 (80.2%)
Cache DeepSeek V4 $0.30/$0.50 per 1M, 90% cache discount
Edge Cloudflare Workers AI Global CDN

Best for: Real-time apps, trading bots, live coding assistants

💰 Cheapest Pro Stack (<$10/month)

Layer Tool Cost
IDE Trae Pro $10/mo (600 fast, DeepSeek V4/GPT-5.4)
API OpenRouter $10 1K req/day + BYOK 1M/month free
CLI Gemini CLI v0.37.1 (Gemini 3.1 Pro/Flash)
Local Ollama Free
Embeddings Jina AI Free tier

Total Cost: ~$10/month for pro-grade everything

🔒 Local Privacy Stack (100% Offline)

Layer Tool Privacy
Models Ollama + Llama 3.3 / Qwen3-Coder Runs locally
IDE Continue.dev + VS Code BYO local models
CLI Aider + local Ollama Git-integrated, offline
Chat UI Open WebUI Self-hosted ChatGPT alternative
Vector DB ChromaDB / LanceDB Local embeddings storage
Speech Whisper (local) Offline transcription

Best for: Healthcare, legal, finance - any sensitive data

🤖 Agentic AI Stack (Autonomous Workflows)

Component Tool Role
Orchestrator n8n / Gumloop Workflow automation
Reasoning DeepSeek R1 / DeepSeek V4 Complex decision making
Execution Qwen3.6-Plus Code generation
Memory ChromaDB / Supabase Vector Long-term context
Embeddings Jina Embeddings v3 (1M tokens/day free) Semantic search
Monitoring LangSmith Trace agent steps

Best for: Autonomous research assistants, code review bots, data processing pipelines

📊 RAG Stack (Document Q&A)

Component Tool Purpose
Framework LlamaIndex / LangChain RAG orchestration
Vector DB ChromaDB / Weaviate / Supabase Document storage
Embeddings E5-Mistral-7B (best accuracy) Text vectorization
Chunking LlamaIndex Smart document splitting
Reranking Cohere Rerank Improve retrieval accuracy
LLM Claude Sonnet 4.6 (79.3%) / GPT-5.4 Answer generation
Eval RAGAS Measure RAG performance

Best for: ExamAi, legal document analysis, knowledge bases

Free LLM API Providers

Fully Free Providers

OpenRouter

Limits: 20 RPM, 29 free models (262K context max, March 2026), models share quota

Google AI Studio

Data is used for training when used outside UK/CH/EEA/EU.

Rate limits: Tier 1 (default): 250 RPD | Tier 2: Requires $250 spend + 30 days

Model Free Tier Limits
Gemini 3.1 Pro [verify: now paid] 250 RPD (Tier 1)
Gemini 3 Flash 1,500 RPD
All others Check console

Note: Data training outside UK/CH/EEA/EU still applies.

NVIDIA NIM

Phone number verification required. Models tend to be context window limited.

Limits: 1K credits signup, up to 5K total, 40 RPM (phone verify required)

  • 46+ models including Llama 3.3 70B, Llama 4 Scout, Mistral Large, Qwen3 235B

Mistral (La Plateforme)

Free tier requires opting into data training; phone verification required

Limits (per-model): 1 req/s, 500K tokens/min, 1B tokens/month

  • Open and Proprietary Mistral models (Mistral Large 3, Small 3.1, etc.)

Mistral (Codestral)

Limits: 30 RPM, 2K RPD confirmed free

  • Codestral (monthly subscription-based, currently free)

HuggingFace Inference Providers

Serverless Inference limited to models <10GB (some popular models >10GB supported).

Limits: ~$0.10/month in credits

  • Various open models across supported providers

Vercel AI Gateway

Routes to various supported providers.

Limits: $5/month

OpenCode Zen

AI gateway with curated models. Free models may use data for improvement.

  • Big Pickle Stealth (S+, 72.0% SWE-bench)
  • MiniMax M2.5 Free (S+, 80.2% SWE-bench)
  • MiMo V2 Pro/Omni/Flash Free
  • Nemotron 3 Super Free
  • GPT 5 Nano
  • Trinity Large Preview Free

Cerebras

Model Limits
GPT-OSS 120B 30 req/min, 60K tokens/min, 900 req/hour, 1M tokens/day
Llama 3.1 8B Same limits as above
Qwen3-235B Available via API

Groq

Model Limits
Llama 3.1 8B 14,400 req/day, 6K tokens/min
Llama 3.3 70B 1,000 req/day, 12K tokens/min
Llama 4 Maverick/Scout 1,000 req/day
Whisper Large v3/v3 Turbo 7,200 audio-sec/min, 2,000 req/day
Qwen3-32B 1,000 req/day, 6K tokens/min
Kimi K2 Instruct 1,000 req/day, 10K tokens/min
GPT-OSS 20B/120B 1,000 req/day, 8K tokens/min
And 15+ more

Cohere

Limits: 20 RPM, 1K req/month (non-commercial only)

  • Command R+ 2026
  • c4ai-aya-expanse/vision-32b
  • command-a/r/r7b variants

GitHub Models

Extremely restrictive input/output token limits.

Limits: Dependent on Copilot subscription tier (Free/Pro/Pro+/Business/Enterprise)

  • AI21 Jamba 1.5 Large
  • Codestral 25.01
  • Cohere Command A, Command R/R+ 08-2024
  • DeepSeek-R1, DeepSeek-R1-0528, DeepSeek-V3.2, DeepSeek-V3-0324
  • Grok 3, Grok 3 Mini
  • Llama 4 Maverick 17B 128E Instruct FP8, Llama 4 Scout 17B 16E Instruct
  • Llama-3.2-11B/90B-Vision-Instruct, Llama-3.3-70B-Instruct
  • MAI-DS-R1, Meta-Llama-3.1-405B/8B-Instruct
  • Ministral 3B, Mistral Medium 3 (25.05), Mistral Small 3.1
  • OpenAI GPT-4.1/mini/nano, GPT-4o/mini, GPT-5/mini/nano
  • OpenAI gpt-5-chat (preview), o1/o1-mini/o1-preview, o3/o3-mini, o4-mini
  • OpenAI Text Embedding 3 (large/small)
  • Phi-4, Phi-4-mini-instruct/reasoning, Phi-4-multimodal-instruct, Phi-4-reasoning

Cloudflare Workers AI

Limits: 10,000 neurons/day

  • @cf/aisingapore/gemma-sea-lion-v4-27b-it
  • @cf/ibm-granite/granite-4.0-h-micro
  • @cf/openai/gpt-oss-120b, @cf/openai/gpt-oss-20b
  • @cf/qwen/qwen3-30b-a3b-fp8
  • @cf/zai-org/glm-4.7-flash
  • DeepSeek R1 Distill Qwen 32B
  • Deepseek Coder 6.7B Base/Instruct (AWQ)
  • Deepseek Math 7B Instruct
  • Gemma 2B/3 12B/7B Instruct (LoRA)
  • Hermes 2 Pro Mistral 7B
  • Llama 2 7B/13B Chat (FP16/INT8/AWQ/LoRA)
  • Llama 3 8B Instruct, Llama 3.1 8B Instruct (AWQ/FP8)
  • Llama 3.2 1B/3B/11B Vision Instruct
  • Llama 3.3 70B Instruct (FP8), Llama 4 Scout Instruct
  • Mistral 7B Instruct v0.1/v0.2 (AWQ/LoRA)
  • Mistral Small 3.1 24B Instruct
  • Qwen 1.5 0.5B/1.8B/7B/14B Chat (AWQ)
  • Qwen 2.5 Coder 32B Instruct, Qwen QwQ 32B
  • Phi-2, SQLCoder 7B 2
  • And more...

Providers with Trial Credits

Provider Credits Duration Notes
Fireworks $1 Permanent Various open models
Baseten $30 Permanent Pay by compute time
Nebius $1 Permanent Various open models
Novita $0.50 1 year Various open models
AI21 $10 3 months Jamba family
Upstage $10 3 months Solar Pro/Mini
NLP Cloud $15 Permanent Phone verification required
Alibaba Cloud 1M tokens/model 90 days Qwen models
Modal $5-30/month Monthly Pay by compute time
Inference.net $1 (+$25 on survey) Permanent Various open models
Hyperbolic $1 Permanent DeepSeek, Llama, Qwen, GPT-OSS
SambaNova Cloud $5 3 months Llama, Qwen, DeepSeek
Scaleway 1M tokens Permanent DeepSeek, Llama, Mistral, Gemma

Additional Free API Providers

Provider Models Free Tier Environment Variable
Together AI 19 Credits/promos vary by account TOGETHER_API_KEY
iFlow 11 Free for individuals (7-day key expiry) IFLOW_API_KEY
ZAI 7 Free tier (generous quota) ZAI_API_KEY
SiliconFlow 6 1K RPM, 50K TPM SILICONFLOW_API_KEY
Perplexity API 4 ~50 RPM default PERPLEXITY_API_KEY
OVHcloud AI Endpoints 8 2 req/min (no key), 400 RPM with key OVH_AI_ENDPOINTS_ACCESS_TOKEN
Chutes AI 4 Free community GPU-powered CHUTES_API_KEY
DeepInfra 4 200 concurrent requests DEEPINFRA_API_KEY
Replicate 2 6 req/min (no payment), up to 3K RPM with payment REPLICATE_API_TOKEN

AI-Powered IDEs

Full-featured integrated development environments with built-in AI assistance.

IDEs with Pro-Grade Models

Cursor

Model: GPT-5.1-Codex-Max (77.9% SWE-bench Verified) [verify]

  • Free tier: 500 slow premium req/mo, 2K completions/mo (post-Dec 2025 credits)
  • Free models: Cursor Small, Deepseek v3, Gemini 2.5 Flash, GPT-4o mini (500/day limit), Grok 3 Mini Beta [verify: GPT-5.4 now paid-only]
  • Claude models removed from free tier ~June 2025
  • Free tier uses token-based usage tracking (not request-based)
  • AI-powered code editor with autonomous coding capabilities
  • Pro ($20/mo or $16/mo annually): Extended Agent limits + Unlimited Tab completions + Background Agents + Maximum context windows
  • Pro+ ($60/mo): 3x usage on all OpenAI, Claude, Gemini models
  • Ultra ($200/mo): 20x usage on all models + Priority access to new features
  • Teams ($40/user/mo): Pro features + Centralized billing + Usage analytics + SAML/OIDC SSO
  • Enterprise (Custom): Everything in Teams + Pooled usage + SCIM + AI code tracking API + Audit logs

Pricing | GPT-5.1-Codex-Max Announcement

Trae

Models: DeepSeek V4, GPT-4.1, GPT-4o, Gemini 2.5 Pro (Claude models removed Nov 2025)

  • 10 fast requests + 50 slow requests/month for premium models
  • 1,000 slow requests/month for advanced models
  • 5,000 auto-completions/month
  • VS Code-based IDE with AI integration
  • No credit card required for free tier
  • Pro ($10/mo): 600 fast + unlimited slow requests for premium models
  • Unlimited slow requests for advanced models
  • Zero rate limits and faster access to premium models
  • Extra packages available: $3-$12 for additional fast requests
  • First month available for $3

Pricing | Documentation

Windsurf

Models: OpenAI, Anthropic, Google, xAI model access

  • 25 prompt credits/month limit
  • Multiple providers (OpenAI, Claude, Gemini, xAI)
  • Credit card required
  • Can purchase add-on credits to continue
  • Pro ($15/mo): 500 prompt credits/month
  • Teams ($30/user/mo): 500 prompt credits/user/month
  • Enterprise ($60+/user/mo): 1,000 prompt credits/user/month

Pricing

Void IDE

Models: Multi-agent (frontend/backend/testing agents)

  • Agent-first IDE - new 2026 category
  • Multiple specialized agents coordinate across codebase
  • Free preview tier with high usage limits
  • VS Code-based

Best for: Full-stack development with natural language direction

Qoder

Models: Qwen3.6-Plus (71.2% SWE), Qwen-Coder-Qoder, GPT-4o, Claude Sonnet [verify: flagship models now paid-only]

  • Free tier: Unlimited completions + limited chat/agent (basic models) + 2-week Pro trial (1,000 credits)
  • Experts Mode: Multi-agent collaboration (new Mar 2026)
  • Quest Mode: Fully autonomous app building
  • Nextnew: Tab predictions
  • Windows/macOS, VS Code-based

Pricing (50% discount - Apr 2026):

  • Free: Basic models, limited messages
  • Pro: $10/mo (reg $20) - 2,000 credits
  • Pro+: $30/mo (reg $60) - 6,000 credits
  • Ultra: $100/mo (reg $200)
  • Credits: $0.01/credit (reg $0.02), expire 1mo

Docs | Pricing

RooCode

Models: Bring your own API keys (any provider)

  • Open-source AI-powered coding assistant for VS Code
  • Whole dev team of AI agents in your editor
  • No subscription required - pay-as-you-go with your own keys
  • Custom modes for different coding tasks

GitHub | Website

IDEs with Basic Models

Codeium

Model: Base model (Llama 3.1 70B), pro-grade models require subscription

  • Individual plan: Free forever with unlimited code completions, AI chat, commands
  • 70+ programming languages supported
  • IDE integrations: VS Code, JetBrains, Vim/Neovim, Jupyter
  • No credit card required
  • Limited context awareness (expanded in paid tiers)
  • Pro ($10/mo): Unlimited usage with advanced context awareness, Claude 3.5 Sonnet, GPT-4o access
  • Teams ($12/user/mo): Pro features + team management
  • Enterprise (Custom): On-premise deployment, custom models

Pricing | Documentation

JetBrains AI Assistant

Models: Local models + cloud models with limited quota

  • AI Free tier included with IDEs
  • Unlimited code completion and local model support
  • Limited quota for cloud-based features
  • 30-day AI Pro trial included
  • Offline mode with local models via Ollama/LM Studio
  • AI Pro ($15/mo): Increased cloud quota + unlimited local models
  • AI Ultimate ($25/mo): Maximum cloud quota + advanced features

AI Pricing | AI Features

Tabnine

Models: Claude 3.5 Sonnet, GPT-4o, Llama 3.3 70B, proprietary models

  • Free tier with limited features
  • Basic AI code completions and chat (limited)
  • Local processing available
  • Context heavily limited in free tier
  • 600+ programming languages supported
  • Pro ($12/mo): Enhanced AI completions and chat
  • Enterprise ($39/user/mo): Multiple LLMs, private deployment, on-premises and air-gapped options

Pricing

SuperMaven ⚠️ DISCONTINUED

Status: Shut down November 21, 2025 after acquisition by Cursor (Nov 2024)

Models: GPT-4o, Claude 3.5 Sonnet, GPT-4 (via chat interface) - Free tier with basic features - Basic code suggestions - 7-day data retention limit - Credit card required for registration - 1M token context window

Historical Note: SuperMaven was acquired by Cursor in November 2024 and officially shut down in November 2025. Features were integrated into Cursor Tab. Users should migrate to Cursor or alternatives.

Pricing

Bolt.new

Models: Unspecified models

  • $1 credit/mo = ~100K tokens (reduced Mar 2026)
  • Specific model not publicly specified
  • Credit card required
  • $20/mo: 20M tokens/month
  • $200/mo: 200M tokens/month

Token Documentation

Lovable

Models: Unspecified models

  • 5 daily credits, max 30 per month (free)
  • Models not publicly enumerated
  • Credit card required
  • Pro ($25/mo): 150 credits/month (5 daily credits)
  • Teams ($30/mo): Higher limits (undisclosed)

Messaging Limits

v0.dev

Models: Proprietary models (not frontier)

  • $5 in credits/month limit
  • Uses proprietary models with varied routing
  • Credit card required
  • GPT-5 access requires v0 Premium subscription

Updated Pricing Blog

Additional 2026 AI Chat Platforms

General-purpose chat interfaces with free tiers.

Platform Free Model Key Capabilities Limitations
ChatGPT GPT-4o / GPT-5.4-limited [verify] Sora 3, DALL-E 4, GPT Store ~20 msgs/3hr
Gemini Gemini 3.1 Flash 2M Context, 20 Deep Research/mo Research quota
Claude Claude Sonnet/Haiku [verify: Opus paid-only] Technical reasoning ~30 msgs/5h
Grok Grok 4.2 Aurora 2 images, voice 15 msgs/12hr
Mistral Le Chat Mistral Medium 3 Structured output Fewer integrations

Notes:

  • Aurora - xAI's image generation model (available in Grok)
  • Sora 2 - OpenAI's video generation (integrated in ChatGPT)
  • DALL-E 4 - OpenAI's latest image model (ChatGPT)
  • Deep Research - Gemini's agentic research feature

CLI Coding Tools

Command-line tools for AI-assisted coding in your terminal.

CLI Tools with Pro-Grade Models

Gemini CLI

Models: Gemini 3.1 Flash [verify: Pro now paid], Gemini 2.5 Pro

  • Gemini 3.1 Pro latest version (v0.37.1 April 2026)
  • 100 requests/day for Gemini 2.5 Pro (free tier fallback)
  • 250 requests/day for Gemini 2.5 Flash
  • No credit card required for free tier
  • MCP server support, Google Search grounding
  • Enable via /settings → Preview features → true
  • Install: npm install -g @google/gemini-cli

Rate Limits | Pricing | Gemini 3 Pro Announcement

Rovo Dev CLI

Important

Rovo Dev CLI isn’t available during a Rovo Dev Standard trial. To use this feature, you need a paid Rovo Dev Standard subscription.

Models: Claude Sonnet 4 [verify], GPT-5 preview [verify]

  • 5M tokens/day free tier
  • No credit card required during beta
  • Token limits reset at midnight UTC
  • Jira/Confluence integration, MCP server support
  • Requires Atlassian account
  • Pro ($19.99/mo via Google AI Pro): 100 tasks/day, 5x higher limits, 5x concurrent tasks (15)
  • Ultra (via Google AI Ultra): 300 tasks/day, 20x higher limits, 60 concurrent tasks, priority access to latest models

Documentation | Token Limits

Warp

Models: GPT-4.1, Claude Opus 4.1 [verify], Claude Sonnet 4 [verify], Gemini 2.5 Pro

  • 150 AI credits/month (first 2 months), then 75 AI credits/month
  • No credit card required for basic signup
  • AI-powered terminal with code generation
  • Build ($20/mo): 1,500 AI credits/month
  • Reload Credits available (up to 50% cheaper than old overage rates, roll over for 12 months)
  • Bring Your Own API Key (BYOK) option available
  • New pricing effective immediately for new customers (Oct 30, 2025)
  • Existing monthly subscribers transition on first renewal after Dec 1, 2025

Pricing

GitHub Copilot

Models: GPT-4.1, Claude Opus 3.5, Gemini 2.0 Flash, Grok Code Fast 1 (Free tier); GPT-5.1-Codex-Max available in Pro/Pro+/Business/Enterprise only

  • 50 agent mode or chat requests + 2,000 completions/month (Free tier)
  • Agent Mode with autonomous multi-step coding
  • No credit card required
  • Free Copilot Pro for students/educators (GitHub Student Pack, Copilot Pro for teachers/maintainers)
  • Limited to basic features after quota
  • Pro ($10/mo): 300 premium requests + unlimited completions/month
  • Pro+ ($39/mo): 1,500 premium requests + unlimited completions/month
  • Business ($19/user/mo): 300 premium requests/user + unlimited completions
  • Enterprise ($39/user/mo): 1,000 premium requests/user + unlimited completions
  • GPT-5.1-Codex-Max available in public preview (Dec 4, 2025) for Pro, Pro+, Business, Enterprise - NOT in free tier
  • Overage billing available at $0.04/request

Plans Details | Agent Mode | GPT-5.1-Codex-Max Preview

Jules

Model: Gemini 2.5 Pro

  • 15 tasks/day free tier
  • 3 concurrent tasks
  • No credit card required
  • Gmail account required (18+ years)
  • Task limits reset on rolling 24-hour window
  • Pro ($19.99/mo): 100 tasks/day, 5x higher limits, 5x concurrent tasks (15)
  • Ultra (via Google AI Ultra): 300 tasks/day, 20x higher limits, 60 concurrent tasks, priority access to latest models

Usage Limits | Documentation | Google AI Plans

AWS Kiro

Models: Claude 4 Sonnet, Claude 3.7 Sonnet (AWS-hosted)

  • 50 credits/month (Free tier)
  • 14-day welcome bonus: 500 credits
  • No credit card required
  • Pro ($20/mo): 1,000 credits
  • Pro+ ($40/mo): 2,000 credits
  • Power ($200/mo): 10,000 credits

Pricing | Introduction Blog

Amazon Q Developer

Model: Claude Sonnet 4 [verify] (AWS-hosted)

  • 50 agentic requests/month limit (multi-turn conversations)
  • Latest Claude models
  • Credit card required
  • Must upgrade to Pro for continued access
  • Perpetual free tier
  • Pro ($19/mo): Increased limits for agentic requests
  • Usage may be adjusted based on regional factors and usage patterns

Pricing

OpenCode

Models: 300+ via OpenRouter (Claude, GPT, DeepSeek, Gemini, Grok, etc.)

  • Open-source AI coding agent (Go-based CLI)
  • Zen Free tier with 8 exclusive models (Big Pickle, MiniMax M2.5 Free, MiMo V2)
  • Privacy-sensitive: no code/context stored
  • opencode run --dangerously-skip-perm for quick execution

GitHub | Website

ForgeCode

Models: 300+ models via OpenRouter (Claude, GPT, O Series, Grok, DeepSeek, Gemini)

  • AI-enabled pair programmer (Rust-based, Apache 2.0)
  • Model-agnostic agent harness
  • Semantic codebase search via :sync
  • 10K tokens/day free tier

GitHub | Website

OhMyPi

Models: Bring your own keys (any provider)

  • AI coding agent for the terminal (Zig-powered)
  • Hash-anchored edits, optimized tool harness
  • LSP integration, Python support, browser automation
  • Subagents with coordinated API rate limiting
  • Multiplexer integration (tmux, GNU Screen, Zellij)
  • Interrupt anytime workflow

GitHub

Goose

Models: Any LLM (Claude, GPT, DeepSeek, etc.)

  • Open-source extensible AI agent from Block (now AAIF/Linux Foundation)
  • Desktop app, CLI, and API
  • Active engineering tasks (not just code suggestions)
  • Built for code, workflows, and automation
  • Model-agnostic architecture

GitHub | Website

Kilo Code

Models: Bring your own API keys (Claude, Gemini, GPT, etc.)

  • Up to $25 signup credits (one-time bonus)
  • Open source VS Code extension
  • Pay-as-you-go with no markup on model pricing
  • Credit card required to claim full bonus credits
  • Full BYOK support

GitHub | Documentation | Pricing

RooCode

Models: Bring your own API keys (any provider)

  • Open-source AI-powered coding assistant for VS Code
  • Whole dev team of AI agents in your editor
  • No subscription required - pay-as-you-go with your own keys
  • Custom modes for different coding tasks
  • Previously known as Roo Cline

GitHub | Website

CLI Tools with Basic Models

Claude Code

Models: Claude Sonnet 4 [verify], Opus 4.5 [verify: paid-only], Haiku 4.5

  • Free tier available with limited usage
  • Pro ($20/mo or $17/mo annually): Sonnet 4 access with more usage
  • Max 5x ($100/mo): ~225 messages/5 hours
  • Max 20x ($200/mo): ~900 messages/5 hours
  • Extended thinking modes: "think" (~4K tokens), "megathink" (~10K), "ultrathink" (~32K)
  • Usage limits reset weekly with 5-hour rolling windows

Pricing | Claude Code Guide

OpenAI Codex CLI

Model: GPT-5.1-Codex-Max (77.9% SWE-bench Verified)

  • Free with ChatGPT Plus ($20/mo): 30–150 messages/5 hours
  • ChatGPT Pro ($200/mo): 300–1,500 messages/5 hours
  • Pay-as-you-go API: $1.25/$10 per million tokens (input/output)
  • Free OSS mode: Access to open-source models only (via --oss flag)
  • First model with "compaction" for multi-million token sessions (24+ hour tasks)
  • 30% fewer thinking tokens than previous GPT-5.1-Codex
  • Cross-platform: macOS 12+, Ubuntu 20.04+, Windows 11 via WSL2

GitHub Repo | GPT-5.1-Codex-Max Announcement

QuantFlow Pilot

Models: Uses Claude Code for implementation

  • Autonomous AI development pipeline — #1 Terminal Benchmark 2.0
  • Turns GitHub issues into pull requests automatically
  • Label an issue "pilot" → Pilot claims it → Creates branch → Plans → Implements → Quality gates → Opens PR
  • Telegram bot integration available
  • Desktop app available
  • Install: brew install qf-studio/tap/pilot or go install github.com/qf-studio/pilot@latest

GitHub | Website

MemoryPalace

Models: Works with any LLM (Claude, ChatGPT, Cursor, Gemini, local models)

  • AI memory system with highest LongMemEval score ever (96.6%)
  • Uses ancient "memory palace" technique for AI conversations
  • Stores conversations in structured format: wings (people/projects), halls (memory types), rooms (specific ideas)
  • Raw verbatim storage without AI summarization
  • Three mining modes: projects (code/docs), convos (conversation exports), general (auto-classified)
  • MCP server with 19 tools for AI integration
  • Local, open, adaptable — runs entirely on your machine
  • Install: pip install mempalace

GitHub | Note from Milla

Continue.dev

Models: Bring your own API keys (200+ models supported)

  • Free VS Code and JetBrains extension
  • Full support for local models via Ollama/LM Studio
  • Solo tier: Private/team/public visibility options
  • Community hub for custom AI assistants
  • No vendor lock-in or usage limits for local models

GitHub | Model Hub

Aider

Models: Bring your own API keys (supports many providers)

  • Free command-line assistant with built-in Git integration
  • Works with GPT-4o, Claude Sonnet, DeepSeek, and local models
  • Multi-file editing with repository context
  • Voice-to-code support
  • Use /help to see all commands

Documentation | GitHub

API Providers for AI Coding Tools

These services provide API access to coding-optimized models for tools like Cursor, Continue.dev, Cline, etc.

OpenRouter

  • 50 requests/day free tier (1,000/day with $10+ credits)
  • Qwen3-Coder-480B, Qwen3-30B-A3B, Qwen3-235B-A22B, Gemini Flash
  • 20 req/min rate limit for free tier
  • OpenAI-compatible API

Cerebras

  • 1.5M tokens/day free tier (expanded Feb 2026)
  • 30 req/min, 8,192 token context
  • Models: Qwen3.6-Plus-480B, Llama 3.1 70B
  • Ultra-fast: 2,400 t/s (Qwen3.6)
  • OpenAI-compatible API (works with Cursor, Continue.dev, Cline, RooCode, etc.)
  • Paid tiers: Developer ($10+ self-serve), Enterprise (custom pricing)

Pricing | API Docs | Integrations

Paid Tiers Comparison

AI-Powered IDEs - Paid Plans

IDE Entry Tier Credits/Requests Key Features
Cursor Pro ($20/mo) Extended Agent limits Unlimited completions
Trae Pro ($10/mo) 600 fast + unlimited slow Zero rate limits
Windsurf Pro ($20/mo) 500 prompt credits Multi-provider
Qoder Pro ($10/mo - 50% off) 2,000 credits Quest Mode, Experts Mode
Codeium Pro ($10/mo) Unlimited Claude 4.6 [verify], GPT-5.4 [verify]
Tabnine Pro ($12/mo) Enhanced completions 600+ languages
JetBrains AI AI Pro ($15/mo) Increased cloud quota Unlimited local models
SuperMaven DISCONTINUED Nov 2025 1M token context Acquired by Cursor

CLI Tools - Paid Plans

Tool Entry Tier Credits/Requests Key Features
Claude Code Pro ($20/mo) ~225 messages/5h Sonnet access [verify]
Warp Build ($20/mo) 1,500 credits/month BYOK available
GitHub Copilot Pro ($10/mo) 300 premium req/month Unlimited completions
Rovo Dev CLI Jira Standard ($7.53/mo) 20M tokens/day 4x free tier
Jules Pro ($19.99/mo) 100 tasks/day 5x free limits
OpenAI Codex CLI ChatGPT Plus ($20/mo) 30-150 msg/5h GPT-5.1-Codex-Max
Amazon Q Developer Pro ($19/mo) Increased agentic limits AWS-hosted Claude
Kilo Code Pay-as-you-go Up to $25 signup credits No markup on models

Local Models

Running open-weight frontier models locally provides unlimited coding assistance without API costs.

Popular Tools:

  • Cline - VS Code extension with Plan/Act modes and MCP support
  • Aider - Command-line assistant with Git integration
  • Continue.dev - Open-source VS Code extension (200+ models)

Local Model Tools:

  • Ollama - Run frontier models locally
  • LM Studio - Easy desktop app for local LLMs (no terminal required)

Notable Local Models (2026):

  • Qwen3.6-Plus-480B (71.2% SWE, ~150GB VRAM)
  • Gemma 4 [verify] (Google, Apache 2.0, fully open-source)
  • GLM-5.1 / GLM-5V-Turbo [verify] (Zhipu MoE-based SOTA coders)
  • Devstral 2 (24B, Apache 2.0, agent-optimized)
  • DeepSeek Coder V4 (lite version ~18GB)
  • Codestral 2 (Mistral, 22B)
  • GLM-4.9-Air (Chinese/English coding)

Note: Frontier models require substantial RAM/VRAM. See Unsloth Qwen3-Coder guide for details.

Update April 2026: Gemma 4 and GLM-5.1 families are new flagship open-source releases. Verify availability in Ollama/LM Studio before downloading.

free-coding-models CLI

Find the fastest free coding model in seconds. Ping 238 models across 25 providers in real-time.

npm install -g free-coding-models
free-coding-models

Features

  • Parallel pings — all 238 models tested simultaneously
  • Stability Score (0-100) — composite score from p95 latency, jitter, spike rate, uptime
  • Smart ranking — top 3 highlighted 🥇🥈🥉
  • Favorites — star models with F, persisted across sessions
  • Tool Integration — auto-configure OpenCode, Goose, Aider, Continue, Cline, etc.
  • OpenCode Zen Models — 8 exclusive free models (Big Pickle, MiniMax M2.5 Free, MiMo V2, etc.)

Quick Usage

# Most reliable model right now
free-coding-models --fiable

# Configure Goose with S-tier model
free-coding-models --goose --tier S

# NVIDIA top models only
free-coding-models --origin nvidia --tier S

# JSON output for scripting
free-coding-models --tier S --json | jq -r '.[0].modelId'

Tool Launcher Flags

Flag Launches
--opencode 📦 OpenCode CLI
--openclaw 🦞 OpenClaw
--goose 🪿 Goose
--aider 🛠 Aider
--qwen 🐉 Qwen Code
--continue ▶️ Continue CLI
--cline 🧠 Cline
--gemini ♊ Gemini CLI
--rovo 🦘 Rovo Dev CLI
And 8 more...

Tier Scale

Tier SWE-bench Best For
S+ ≥75% Claude Opus 4.6 [verify], GPT-5.4 [verify]
S 65-75% Qwen3.6-Plus (71.2%), Claude Sonnet 4.6 [verify]
A+/A 40–60% Solid alternatives
A-/B+ 30–40% Smaller tasks
B/C < 30% Code completion

License Summary

All 238 models allow commercial use of generated output. You own what the models generate.

License Models Commercial
Apache 2.0 Qwen3/Qwen2.5 Coder, GPT-OSS 120B/20B, Devstral Small 2, Gemma 4, MiMo V2 Flash ✅ Unrestricted
MIT GLM 4.5/4.6/4.7/5, MiniMax M2.1, Devstral 2 ✅ Unrestricted
Llama Community License Llama 3.3 70B, Llama 4 Scout/Maverick ✅ Attribution required. >700M MAU → separate Meta license
DeepSeek License DeepSeek V3/V3.1/V3.2, R1 ✅ Use restrictions on model (no military, no harm) — output is yours
NVIDIA Nemotron License Nemotron Super/Ultra/Nano ✅ Updated Mar 2026, now near-Apache 2.0 permissive
MiniMax Model License MiniMax M2, M2.5 ✅ Royalty-free, non-exclusive. Prohibited uses policy applies to model
Proprietary (API) Claude (Rovo), Gemini (CLI), Perplexity Sonar, Mistral Large, Codestral ✅ You own outputs per provider ToS
OpenCode Zen Big Pickle, MiMo V2 Pro/Flash/Omni Free, GPT 5 Nano, MiniMax M2.5 Free, Nemotron 3 Super Free ✅ Per OpenCode Zen ToS

Key Points:

  1. Generated code is yours — no model claims ownership of your output
  2. Apache 2.0 / MIT models (Qwen, GLM, GPT-OSS, MiMo, Devstral Small) are the most permissive — no strings attached
  3. Llama requires "Built with Llama" attribution; >700M MAU needs a Meta license
  4. DeepSeek / MiniMax have use-restriction policies (no military use) that govern the model, not your generated code
  5. API-served models (Claude, Gemini, Perplexity) grant full output ownership under their terms of service

⚠️ Disclaimer: This is a summary, not legal advice. License terms can change. Always verify the current license on the model's official page before making legal decisions.

Comparison Notes

  • Goal: Compare AI coding tools by their access to pro-grade models and free tier limits.
  • What qualifies a model as "pro-grade"? Models must achieve ≥60% on SWE-bench Verified, demonstrating real-world software engineering capability. Current qualifying models: Claude Opus 4.5 (80.9% [verify]), GPT-5.1-Codex-Max (77.9% [verify]), Claude Sonnet 4.5 (77.2% [verify]), Gemini 3 Pro (76.2% [verify]), GPT-5 (74.9% [verify]), Claude Opus 4.1 (74.5% [verify]), Claude Sonnet 4 (72.7% [verify]), GPT-5 mini (71.0% [verify]), Qwen3-Coder-480B (69.6% [verify]), and Gemini 2.5 Pro (63.2% [verify]).
  • [verify] tag: Indicates information needs verification from official sources. Pricing, limits, and model availability change frequently.
  • Different limit types: Tools use various quota systems - requests, tokens, credits, chats - making direct comparison challenging. Check documentation for specifics.
  • Real-world usage: Actual consumption varies dramatically based on coding style, task complexity, and tool implementation.

Education & Student Programs

Program What You Get Requirements
GitHub Student Pack Free Copilot Pro for students Verify with .edu email
GitHub Copilot Free 50 chat + 2,000 completions/month VS Code users
Copilot Pro for Teachers/Maintainers Free Copilot Pro Open source maintainers & educators

Additional 2026 AI Tools

Agentic Workflow Platforms

Visual orchestration tools for building autonomous AI agents without coding.

Platform Free Tier Best For Key Features
Make (Integromat) 1,000 ops/month Visual builders Drag-and-drop AI Agents, 3,000+ app integrations
n8n Unlimited (self-hosted) Technical teams Self-hosted RAG systems, private data automation
Gumloop 2,000 credits/month No-code agents Natural-language builder, "Gummie" troubleshooting agent
Relay.app Generous free plan Beginners Simple agentic workflows
Activepieces 1,000 tasks/month Open-source Flat pricing, self-hostable
Podium Entry-level tiers Sales/communication 24/7 lead response AI agents
QuantFlow Pilot Free Autonomous development #1 Terminal Benchmark 2.0 — AI that ships your tickets

Data Visualization & Analysis

AI-powered tools for conversational data analysis and narrative visualization.

Tool Function Free Tier Detail Key Feature
Julius Chat-with-data Upload spreadsheets, generate instant visualizations
Anomaly AI AI Dashboards Generate interactive dashboards from natural language
Flourish Data Storytelling No-code interactive maps, "scrollytelling" features
Datawrapper Publishing Publish-ready charts in seconds, journalism-focused
Looker Studio Marketing Data Seamless Google Analytics/Ads integration
Power BI Desktop Microsoft reports Copilot recommendations, local report building

Creative & Multimedia Tools

Professional-grade content creation with generous free tiers.

Tool Output Free Tier Key Capability
Veo Video Basic Free Cinematic clips with realistic motion and sound
Sora 2 (via ChatGPT) Video Limited free tier Deep ChatGPT integration, high-quality video
DALL-E 4 (via ChatGPT) Image Limited free tier Latest OpenAI image model
Synthesia Video Avatars Free individual plan "Video Agents" in 120+ languages
1 More Shot Music Videos Free plan Advanced lip-sync, frame-by-frame control
Leonardo.Ai Images 150 tokens/day (~70 images) Commercial use allowed
Recraft AI Vector/SVG 30 credits/day Infinitely scalable icons and logos
Ideogram Images 10-20 prompts/day Perfect text rendering, "Magic Prompt"
Suno AI Music 50 credits/day (~10 tracks) Complete songs with vocals and instruments
ElevenLabs Voice Basic Free Realistic voice cloning
Canva AI Design Robust free tier AI design assets, brochures, short videos

Productivity & Research Tools

Tool Function Free Tier Detail Key Feature
Grammarly Writing 100 AI prompts/month Rewrites and tone detection
LanguageTool Grammar 10,000 characters/text 25+ languages, open-source
Fathom Meetings Forever Free Records/transcribes Zoom/Teams, auto-sync to CRM
NotebookLM Research Free Audio Overview podcasts, grounded in your documents
Humata PDF Analysis 60 pages/month Clickable source citations
QuillBot Rewriting 125 words/time Fluency & Standard modes
DeepL Translation Basic Free Incognito sensitive mode
MemoryPalace AI Memory Free, open source 96.6% LongMemEval — memory palace technique for AI

Vertical AI (Specialized Domains)

Medical AI:

Tool Pricing Key Value
iatroX Free Adaptive Q-Bank, NICE/BNF clinical reference
DxGPT Free Diagnostic assistant (500K+ users, 6K doctors)
OpenEvidence Free (US verified) Evidence-grounded search, ambient note generation

Legal AI:

Tool Pricing Key Value
DocLegal.Ai $10/month Clause suggestion, risk detection
Doculex.ai Varies Case-data-driven drafting from medical records
Spellbook 7-day trial In-editor contract analysis
Harvey AI Enterprise Regulatory matters, high security

Marketing & SEO Tools

Tool Function
Wellows AI Visibility Score tracking across ChatGPT, Gemini, Perplexity
Google SGE Labs See how AI Overviews interpret target keywords
NeuronWriter AI content scoring
Surfer SEO Content optimization
Jasper AI copywriting with brand voice
Writesonic Scalable copywriting

Open Source & Local Tools

Tool Function Description
Open WebUI Local Chat Interface ChatGPT-like experience running entirely offline with Ollama
Whisper (OpenAI) Speech-to-Text Most accurate open-source transcription
Piper Text-to-Speech High-quality offline audio generation
ComfyUI Image Generation Node-based interface for Stable Diffusion
Zed AI IDE 50 AI prompts/month, native performance, high speed
Void IDE Agent-first IDE Multi-agent frontend/backend/testing
MemoryPalace AI Memory System 96.6% LongMemEval — memory palace technique for AI conversations

⚡ Realtime & Streaming APIs

Low-latency APIs for voice assistants, live coding copilots, trading tools, and realtime chat.

Streaming LLM APIs

Provider Latency Best For Free Tier
Groq Streaming ~50-150ms (0.4ms/token) Live coding, chat 14.4K req/day
OpenAI Realtime API Low Voice assistants, agents No free tier (pay-per-use only, trial credits new accounts)
Gemini Live API Low Multimodal streaming Dynamic caps (varies by prompt complexity)
Cerebras 2,400 tok/sec (Qwen3.6) Batch + streaming 1.5M tokens/day
Cloudflare Workers AI Edge Global low-latency 10K neurons/day

Speech Streaming APIs

Provider Type Latency Free Tier
Deepgram STT streaming ~300ms $200 credits
AssemblyAI Streaming Realtime STT ~400ms 50 hours/month
Groq Whisper STT fast ~200ms 2,000 req/day
ElevenLabs Streaming TTS streaming ~100ms 10K chars/month
OpenAI Realtime STT + LLM + TTS ~200ms Limited

Best for:

  • Trading bots: Groq streaming (fastest)
  • Voice assistants: OpenAI Realtime API (end-to-end)
  • Live captions: AssemblyAI or Deepgram
  • Realtime chat: Gemini Live API

🎙️ Speech Models

Speech-to-text and text-to-speech models comparison.

Speech-to-Text (STT)

Model Provider Accuracy Speed Free Tier Best For
Whisper Large v3 OpenAI/Groq/Local Excellent Fast 2,000 req/day (Groq) General purpose, local
Deepgram Nova Deepgram Superior Very Fast $200 credits Production, enterprise
AssemblyAI AssemblyAI Excellent Fast 50 hours/month Streaming, diarization
Whisper API OpenAI Excellent Medium Pay-per-use Reliable, consistent
Google Speech Google Cloud Good Fast 60 min/month Google ecosystem
Whisper (local) OpenAI/Ollama Excellent GPU-dependent Unlimited offline Privacy, cost control

Text-to-Speech (TTS)

Model Provider Quality Speed Free Tier Best For
ElevenLabs ElevenLabs 🏆 Best Fast 10K chars/month Voice cloning, pro voice
OpenAI TTS OpenAI Excellent Fast Pay-per-use Reliable, cheap
Piper Local Good Very Fast Unlimited offline Privacy, self-hosted
Bark Suno/Local Good Medium Free (local) Expressive, local
Google TTS Google Cloud Good Fast 1M chars/month Google ecosystem
WhisperSpeech Local Good Fast Unlimited Whisper-based TTS

All-in-One Voice APIs

API Input Output Latency Use Case
OpenAI Realtime Audio Audio ~200ms Voice agents
Deepgram Voice Audio Text/Audio ~300ms Voice bots
AssemblyAI LeMUR Audio LLM response ~1s Voice RAG

🎨 Image Generation Models

Comparison of image generation models and APIs.

Model Provider Quality Speed Free Tier Best For
FLUX.2 Black Forest Labs 🏆 Excellent Fast Local/Replicate High quality, open
DALL-E 4 OpenAI 🏆 Best Medium ChatGPT Plus Latest OpenAI
Ideogram 2.0 Ideogram Excellent Fast 20 prompts/day Text in images
Recraft V4 Recraft Excellent Fast 50 credits/day Vector/SVG output
Stable Diffusion XL Stability AI Good Fast Local/DreamStudio Flexibility, local
Midjourney v6 Midjourney 🏆 Excellent Slow None (paid only) Artistic, Discord
Leonardo.ai Leonardo Very Good Fast 150 tokens/day Commercial use, gaming
Adobe Firefly Adobe Good Fast 25 credits/month Safe, commercial
Imagen 3 Google Excellent Medium Vertex AI trial Photorealistic
DiffusionBee Local Good Fast Local unlimited Easy setup, open-source
ComfyUI Local Good Fast Local unlimited Advanced, node-based

Free Image Model APIs

Provider Model Free Tier Notes
Replicate FLUX.1-schnell Free tier Fast inference
Pollinations Various Unlimited No signup
HuggingFace SDXL/FLUX $0.10 credits Inference API
Leonardo Phoenix 150 tokens/day Commercial OK

🎬 Video Generation APIs

Text-to-video and image-to-video generation. Hot area in 2026.

Model Provider Quality Duration Free Tier Best For
Veo 3 Google 🏆 Excellent 1080p, 60s clips Limited preview Cinematic, realistic
Sora 3 OpenAI 🏆 Excellent 120s ChatGPT Plus High quality, physics
Runway Gen-3 Runway Excellent 10 seconds 3 free credits Creative, filmmaking
Pika 3.0 Pika Very Good 3-5 seconds Free tier Lip-sync improved
Luma Dream Machine Luma Very Good 5 seconds 30 generations/mo Fast, realistic
Kling Kuaishou Excellent 2-10 minutes Limited Long-form, Chinese
Hailuo AI MiniMax Good 6 seconds Free tier Character consistency
Stable Video Diffusion Stability Good 4 seconds Local Open, flexible

Video API Pricing (approximate)

Provider Cost per video Generation time
Runway ~$0.20-0.50 1-5 min
Pika ~$0.10-0.30 30s-2 min
Luma ~$0.30-0.60 2-5 min
Kling ~$0.05-0.20 1-10 min

🌐 AI Browser Automation

Tools for AI agents to control browsers - web scraping, form filling, testing.

Tool Type Pricing Best For
Browserbase Managed browsers $5 free tier Production agents
Steel.dev Browser API Free tier AI-native browser control
Stagehand AI browser framework Open source Next-gen Playwright
Playwright Browser automation Free Reliable, well-documented
Puppeteer Chrome automation Free Chrome-specific
Selenium Cross-browser Free Legacy support
Scrapy Web scraping Free Data extraction

AI-Native Browser Tools

Tool AI Integration Use Case
Stagehand Natural language commands AI agents controlling browsers
Browserbase Session recording for AI Training agent trajectories
Steel.dev Built for LLM agents Agent-native browser API

Stack Recommendation:

  • AI agents: Stagehand + Browserbase
  • Web scraping: Playwright + Scrapy
  • Testing: Playwright + AI assertions

💾 Cheap Vector DB Hosting

Production-ready vector storage without high costs.

Provider Type Free Tier Paid Best For
Supabase Vector Postgres + pgvector 500MB $25/mo starter Full-stack apps
Neon Serverless Postgres 500MB $19/mo Serverless, branching
Railway Managed Postgres $5 credits Usage-based Easy deployment
PlanetScale MySQL + vectors 5GB $39/mo Scale, branching
Chroma Cloud Vector-native Free tier Usage-based Pure vector workloads
Qdrant Cloud Vector DB 1GB $25/mo High performance
Pinecone Managed vector 2GB $70/mo Production, no ops
Weaviate Cloud Vector DB 5M vectors $25/mo Hybrid search
LanceDB Embedded/Cloud Free Cloud beta Multimodal

Self-Hosted (Free Forever)

Database Best For Notes
ChromaDB Prototyping Simple, Python-native
Qdrant Production Rust-based, fast
Milvus Enterprise Scalable, complex
pgvector Postgres apps Just add extension
LanceDB Embedded No server needed

Recommendation by Stage:

  • MVP: ChromaDB (local) → Supabase (hosted)
  • Production: Qdrant Cloud or Pinecone
  • Enterprise: Milvus or Weaviate

🏛️ Common AI Architecture Patterns

Proven patterns for building AI applications.

1. 🤖 Chatbot Architecture

User → Chat UI → LLM API → Response
            ↓
        Context Memory (Redis/Postgres)

Stack:

  • Frontend: Next.js + Vercel AI SDK
  • Backend: FastAPI + OpenRouter
  • Memory: Upstash Redis or Supabase

2. 📚 RAG Architecture (Like ExamAi)

Documents → Chunking → Embeddings → Vector DB
                                    ↓
User Query → Embedding → Similarity Search → LLM → Response

Stack:

  • Framework: LlamaIndex or LangChain
  • Embeddings: BGE-Large or Jina v3
  • Vector DB: ChromaDB (dev) → Pinecone (prod)
  • LLM: Claude Sonnet [verify] or GPT-4o

3. 🎯 Agent Architecture

User Request → Agent Controller → Tool 1 (Search)
                              → Tool 2 (Code exec)
                              → Tool 3 (API call)
                              ↓
                        Synthesize → Response

Stack:

  • Framework: LangGraph, AutoGen, or CrewAI
  • Tools: Function calling with Claude/GPT-4
  • Memory: Vector DB + State management
  • Monitoring: LangSmith or Arize

4. 🔄 Multi-Model Routing Architecture

User Request → Router (classify intent)
                    ↓
    ┌───────────────┼───────────────┐
    ↓               ↓               ↓
Cheap Model    Medium Model    Expensive Model
(GPT-5 Nano)      (Claude Sonnet [verify]) (Claude Opus [verify])
    ↓               ↓               ↓
Simple Q&A    Complex task    Hard reasoning

Implementation:

  • Router: Fine-tuned classifier or LLM-based
  • Cost optimization: Route 80% to cheap models
  • Fallback: Escalate if cheap model fails

5. ⚡ Realtime Streaming Architecture

Audio Input → STT → LLM → TTS → Audio Output
     ↓           ↓      ↓       ↓
 Deepgram    Groq   Claude  ElevenLabs

Stack:

  • STT: Deepgram or Whisper Streaming
  • LLM: Groq for speed or OpenAI Realtime
  • TTS: ElevenLabs or OpenAI TTS
  • Latency target: <500ms end-to-end

6. 🖼️ Multimodal Pipeline Architecture

Image Input → Vision LLM → Structured Output
                                 ↓
                          Database / Action

Stack:

  • Vision: GPT-4o Vision or Gemini 2.5 Pro
  • Structured output: Instructor + Pydantic
  • Storage: Postgres JSONB or MongoDB

7. 🎨 Creative Generation Pipeline

Text Prompt → LLM Enhancement → Image Gen → Upscaling
                                                ↓
                                           Video Gen (optional)

Stack:

  • Enhancement: GPT-4 or Claude
  • Image: FLUX or DALL-E 3
  • Upscale: Upscayl or Magnific
  • Video: Runway or Pika

💵 Model Price Comparison (per 1M Tokens)

API pricing for budget planning. Sorted by input cost.

Model Provider Input Output Cache Hit Best For
MiniMax M2.6 MiniMax $0.08 $0.12 - Bulk generation
DeepSeek V4 DeepSeek $0.28 $0.55 $0.03 🎯 Coding, cached
GLM 4.9 Air ZAI $0.35 $0.75 - Chinese/English
Gemini 3.1 Flash Google $0.30 $0.90 - 2M context
GPT-5 Nano OpenAI $0.45 $1.80 - Cheap reasoning
Qwen3-Coder Alibaba ~$0.60 ~$1.20 - Strong agent tasks
Gemini 2.5 Pro Google $1.25 $10.00 $0.625 High quality, 1M context
GPT-4.1 OpenAI $2.00 $8.00 - General purpose
GPT-5.4 OpenAI $2.50 $10.00 $1.25 Latest OpenAI model
Claude Sonnet 4 Anthropic $3.00 $15.00 $0.60 Best coding, reasoning
Claude Opus 4.6 Anthropic $5.00 $25.00 $2.50 Complex reasoning

💡 Pro tip: DeepSeek's 90% cache discount makes it cheapest for repetitive tasks with long prompts.

🎯 Best Models by Use Case

Don't just use SWE-bench - match models to your specific task.

💻 Coding & Software Engineering

Model Why Free Tier
Claude Sonnet 4.6 79.3% SWE-bench, excellent at following instructions 25 msgs/5h (Claude Code)
Qwen3.6-Plus 71.2% SWE-bench, Chinese + English, agent-optimized 2,000 req/day
GPT-5.4 [verify: paid-only] 80.1% SWE-bench, long context compaction ChatGPT Plus/Pro
DeepSeek V4 Near-Sonnet performance at 1/10th cost DeepSeek API

🧠 Complex Reasoning & Analysis

Model Why Free Tier
DeepSeek R1 Specialized reasoning model, math/logic DeepSeek API
Claude Opus 4.6 84.2% SWE-bench, best for complex architecture Claude Code Pro
Gemini 3.1 Pro 77.4% SWE-bench, 2M context for deep analysis 100 req/day
o3-mini / o1 OpenAI reasoning models, step-by-step ChatGPT Plus

💰 Cheap Bulk Generation

Model Why Cost per 1M
Gemini 2.5 Flash 1M context, high throughput ~$0.35/$1.00
GPT-5 Nano Newest cheap model from OpenAI $0.50/$2.00
GPT-4o ChatGPT free tier model, fast Variable (free tier)
GLM 4.5 Air Good quality, extremely cheap ~$0.40/$0.80
MiniMax M2.7 80.2% SWE-bench, dirt cheap $0.08/$0.12

🤖 Agents & Autonomous Tasks

Model Why Free Tier
Claude Sonnet 4.6 Best tool use, reliable agent behavior Various
GPT-5.4 [verify: paid-only] Compaction for 24+ hour sessions ChatGPT Plus/Pro
Qwen3.6-Plus Built for agentic workflows 2,000 req/day
Big Pickle (OpenCode) 72% SWE-bench [verify], agent-optimized Zen Free tier

👁️ Vision & Multimodal

Model Why Free Tier
Gemini 2.5 Pro Vision 1M token context for images/video 20-100 req/day
GPT-4o Best overall vision capabilities ChatGPT Free
Claude 4 Vision Detailed image analysis Claude Free tier
Qwen2.5 VL Strong open vision model Hyperbolic

🔊 Audio & Speech

Model Provider Free Tier
Whisper Large v3 Groq / Local 2,000 req/day or unlimited local
ElevenLabs ElevenLabs Basic free tier
Piper Local Free, offline TTS

⏱️ Rate Limit Comparison

Critical for scaling applications. Plan your architecture.

Provider RPM TPM Daily Best For
Groq 30 Medium 14,400 High-throughput apps
Cerebras 30 1,000,000 14,400 Batch processing
Gemini Studio 15 High 1,500 Prototyping
OpenRouter 20 Medium 50-1,000 Flexible routing
Cloudflare 300 10K neurons 10K neurons Edge deployment
Groq (varies) 30-50 6K-30K 1K-14.4K Model-dependent

Scaling Strategy by Use Case

App Type Recommended Stack
ExamAi (your app) Cerebras (Qwen3.6-Plus) + Groq
AI Reel Generator Gemini 3.1 Flash (video) + Groq (audio)
Trading AI Groq + local Qwen3.6-Plus
Chatbot OpenRouter + Gemini 3.1 Flash (cheap)
Code Review Bot DeepSeek V4 (cheap) + Claude Sonnet [verify] (quality)

✅ Commercial Use Summary

Quick reference for legal safety.

Provider Commercial Use Notes
OpenRouter ✅ Yes All models
Groq ✅ Yes All models
Gemini API ✅ Yes Per Google ToS
Cohere ✅ Yes 1K req/month free
Claude (API) ✅ Yes Per Anthropic ToS
OpenCode Zen ✅ Yes Per Zen ToS
DeepSeek ✅ Yes No military use restriction
Qwen/Alibaba ✅ Yes Apache 2.0 models
Ollama Local ✅ Yes Fully offline

⚠️ Always verify current ToS - licenses can change.

🧩 RAG Stack Tools

Build document Q&A systems like ExamAi.

Orchestration Frameworks

Tool Best For Free Tier
LlamaIndex Production RAG Open source
LangChain Flexibility Open source
Haystack Enterprise Open source
Vercel AI SDK Edge RAG Free tier

Vector Databases

Database Type Free Tier Best For
ChromaDB Local Unlimited Prototyping, small apps
LanceDB Local/Serverless Generous Multimodal, embeddings
Weaviate Cloud/Local 5M vectors Production scale
Supabase Vector Postgres 500MB Full-stack apps
Pinecone Managed 2GB (1 pod) Production, no ops
Qdrant Local/Cloud 1GB cloud High performance

RAG Evaluation

Tool Purpose
RAGAS Evaluate retrieval quality
LlamaIndex Evals Built-in RAG metrics
Arize Phoenix Observability

🔢 Best Free Embedding APIs

Essential for RAG - don't overlook these.

Embedding Provider Dimensions Free Tier Best For
text-embedding-3-small OpenAI 1536 200K tokens/day General purpose
Jina Embeddings v3 Jina AI 1024 1M tokens/day Multilingual
BGE-Large-EN-v1.5 HuggingFace/Local 1024 Free High quality retrieval
E5-Mistral-7B Various 4096 Varies Best accuracy
Nomic Embed v1.5 Nomic 768 Free tier Long context (8K)
GTE-Large Alibaba 1024 DashScope free Chinese + English

Self-Hosted (Free Forever)

Model Size Speed Quality
BGE-Small 33M Fast Good
MiniLM-L6 22M Very Fast Basic
Nomic Embed 137M Fast Excellent

🖥️ AI Hosting & GPU Providers

Scale beyond free tiers.

Provider Type Pricing Best For
Modal Serverless GPU $5-30/month credits Batch inference
RunPod GPU Cloud $0.20-0.50/hr Training, fine-tuning
Vast.ai Spot GPUs Cheap spot prices Budget inference
Lambda Labs GPU Cloud ~$0.60/hr A100 Stable workloads
Beam.cloud Serverless Per request Spiky traffic
Baseten Model serving $30 credits Production models
Replicate Model hosting 6 req/min free Quick deployment

Serverless Inference (Pay-per-use)

Platform Cold Start Best For
Modal Fast Python functions
Beam Fast ML models
Replicate Medium Pre-built models
HuggingFace Inference Medium HF ecosystem

📊 AI Evaluation Tools

Benchmark your models before production.

Tool Purpose Free Tier
Promptfoo Prompt testing, red-teaming Open source
LangSmith Tracing, evals 5K traces/month
RAGAS RAG evaluation Open source
DeepEval LLM unit testing Open source
Arize Phoenix Observability Generous free tier
Weights & Biases Experiment tracking Academic free

📐 Structured Output Tools

Force LLMs to return valid JSON/schemas.

Tool Approach Best For
Instructor Pydantic validation Python apps
Guidance Constrained generation Complex schemas
Outlines Regex/constrained Fast inference
JSONformer Structure-aware decoding Local models
Zod + Vercel AI SDK TypeScript validation Web apps

🏷️ Legend

Quick reference for badges used in this guide.

Badge Meaning
🟢 No credit card required
💳 Credit card required
⚡ Fast inference (low latency)
🧠 Strong reasoning capabilities
💻 Coding optimized
📦 Open source / self-hostable
🔒 Privacy focused / local
🤖 Agentic capabilities
🎯 Best value / cheap
🌐 Multilingual support
[verify] Needs verification from official source

Contributing

If you spot an error, missing source link, or have updated quota/model information, please open an issue or pull request with a source.

No affiliation with any vendor. All trademarks belong to their owners. Information is for research; accuracy not guaranteed; limits/pricing change frequently.

Related Resources

Research Methodology

This list was compiled and verified using:

  • Gemini - For research and discovering new/additional AI tools
  • Perplexity - For verifying information accuracy and checking if data is current
  • Community repos - All referenced repositories above were used as reference sources

License

MIT © ShaikhWarsi

Last updated: April 11, 2026 • PRs/issues welcome

About

Curated list of free and low cost AI tools, LLM APIs, IDEs, agents, and infrastructure for building real AI apps

FreeAiToolsList.vercel.app

Topics

open-source workflow awesome tools ai awesome-list ai-agents awesome-ai developer-resources ai-tools ai-top-tools theresanaiforthat awesome-ai-tools ai-tools-directory ai-tools-list free-ai-tools ai-directory vibe-coding free-ai-tools-for-pc best-ai-tools-2026

Resources

Readme

License

MIT license

Contributing

Contributing
Activity

Stars

135 stars

Watchers

1 watching

Forks

16 forks
Report repository

Contributors

Languages