AI Agent Tools With a Free Tier
42 of the 64 tools in our index offer a genuine, ongoing free tier — not a time-limited trial or a one-off credit grant — so you can build or test before paying. Developer tools lead by a wide margin; not one of our 7 support-agent listings has one.
We checked every listing in our index against its own transcribed, dated pricing tiers and found 42 of 64 offer a genuine, ongoing way to use the product at zero cost — not a time-limited trial, not a one-off credit grant, but a tier you can stay on indefinitely. The split is not even across categories: coding agents (17 of 19), agent frameworks (11 of 11, unanimous — the newest arrival, Claude Agent SDK, keeps the category perfect) and agent platforms (4 of 10) overwhelmingly let you start free, because these are self-serve, developer-led purchases where the vendor wants a working integration in your hands before you ever talk to sales. Support agents are the mirror image: not one of the site's 7 support-agent listings has an ongoing free plan — every one is sold through a pilot, a per-resolution contract, or a time-limited trial that converts to paid. Voice agents split down the middle (2 of 6): most platforms price per minute from the first call, with only trial credits standing in for a free tier.
We deliberately excluded 5 tools whose pricing page uses free-sounding language that doesn't hold up to that bar. Vapi and Retell AI each disclose, in their own pricing copy, trial-credit-only language (Vapi: 'no permanent free tier'; Retell AI: '$10 in free credits, no commitment') — both are usage-priced per minute from the first call once credits run out. Bland AI waives a separate platform fee on its entry tier but still bills per minute from call one. Intercom Fin and PolyAI each offer a real trial — 14 days and 2 months respectively — that converts to a paid, sales-led plan, not an ongoing free option. None of the 5 are wrong to consider — a free trial is still a real way to test a tool — but they answer a different buyer question than 'can I keep using this for free,' so we left them off this list rather than blur the two.
Why this collection
A tool earns a place here only if its own Published, sourced pricing tiers show an ongoing, indefinite way to use it at $0 — self-hosted and free forever, a permanent free-forever plan, or a cloud tier with a real recurring usage allowance and no stated expiry. A time-limited trial (however generous), a one-time credit grant, or a platform-fee waiver that still bills usage from the first action does not qualify, even where a vendor's own marketing calls it "free." Verified against each listing's dated pricingTiers array, not a vendor's marketing headline; 5 candidates were checked and excluded for this exact reason (see intro) — 2 of the 5 (Vapi, Retell AI, Bland AI) also had a stale attributes.freeTier flag corrected on their own listing as part of this same research pass, since the flag directly contradicted their own already-published pricing copy.
- Cursor
The AI code editor, an agent that reads, writes and edits across your whole codebase.
The Hobby tier is free forever with limited agent usage — enough to try Cursor before choosing a paid plan.
- GitHub Copilot
The mature AI pair-programmer, now with an agent mode, deep in the GitHub + IDE stack.
2,000 code completions and 50 chat requests per month, forever — the most generous free coding-agent tier we found.
- Devin
Cognition’s autonomous AI software engineer, delegate a ticket, get a reviewed pull request back.
A Free tier ships with Devin Review and DeepWiki included, though usage is limited compared to the paid Pro plan.
- Windsurf (now Devin Desktop)
The agentic IDE that was Codeium, then Windsurf, now Cognition’s Devin Desktop, a command centre for coding agents.
The full IDE is free with limited agent usage — a real, ongoing plan, not a time-boxed trial.
- Cline
The open-source coding agent, an autonomous, bring-your-own-key agent that lives in your editor, terminal or SDK.
Free forever and open-source (Apache-2.0) — you only ever pay your own model provider for API usage.
- Aider
AI pair programming in your terminal, an open-source CLI agent that edits your code and commits to git.
Free forever and open-source (Apache-2.0) — bring your own API key, or run a local model for zero cost.
- Replit Agent
Describe an app and Replit Agent builds, tests and deploys it in your browser, a full-stack app factory for anyone.
The Starter tier grants free daily Agent credits and lets you publish one project, ongoing.
- Bolt.new
StackBlitz's prompt-to-app builder that runs a real Node.js environment inside the browser tab itself, no remote VM.
300K daily / 1M monthly tokens free, forever — Bolt branding included on generated sites at this tier.
- Juggler
An open-source, GUI-based workbench for AI coding agents, branching session trees, not a linear transcript. Free, AGPLv3, by JUCE creator Julian Storer.
Free forever, open-source (AGPLv3 + Apache-2.0 SDK) — no signup or account, bring your own model key.
- OpenAI Codex CLI
OpenAI's open-source (Apache-2.0) terminal coding agent, bundled into ChatGPT plans from Free through Pro — Business seats are now grandfathered-only.
The ChatGPT Free plan includes limited Codex access and limited GPT-5.5 Instant access.
- OpenCode
A free, open-source (MIT) AI coding agent for the terminal, desktop and IDE, bring any of 75+ model providers, including your own local Ollama models.
Free, unlimited local use with your own API key across 75+ providers — or run local Ollama models at zero cost.
- Zot
A free, MIT-licensed coding-agent harness in one Go binary, branch any session from any past message and view it as a tree, plus a Telegram bridge.
Free forever, open-source (MIT) — a single Go binary with no signup, bring your own key or run local Ollama models.
- OpenHands
Open source AI agent platform for cloud coding agents, from free local use to enterprise self-hosted deployment.
The open-source core is free (MIT, self-hosted); a free hosted Cloud tier is also available, capped at 1 user and 10 conversations/day.
- CodeRabbit
An AI code-review agent, not a code-writing one — reviews every pull request automatically across GitHub, GitLab, Bitbucket and Azure DevOps, with a genuine free tier for unlimited repos.
Unlimited public and private repos free, forever — plus a 14-day trial of the paid Pro Plus tier layered on top.
- Antigravity CLI
Google's terminal-native coding agent that replaced the open-source Gemini CLI in June 2026 — same Antigravity 2.0 agent core, now closed-source and tied to a Google AI subscription.
The Individual tier is free with unlimited tab completions and command requests, on a weekly-refreshing rate limit.
- Warp
The AI-native terminal that replaces your shell and orchestrates Claude Code, Codex and its own Warp Agent side by side.
150 credits/month free for the first two months, then 75/month ongoing — reduced, but never zero.
- LangChain
The most widely-used framework for building LLM apps and agents, with LangGraph for stateful control.
The core framework is free open-source; LangSmith's Developer tier adds free observability up to 5,000 traces/month.
- CrewAI
A popular open-source Python framework for role-based, collaborating multi-agent teams, free to self-host, now a stable 1.0 release with an optional managed AMP cloud layer.
The framework is free open-source; the hosted AMP Basic tier adds 50 free workflow executions/month.
- Microsoft AutoGen
Microsoft’s open-source framework for orchestrating conversations between multiple AI agents.
Microsoft's framework is free, open-source (MIT) and self-hosted — pay only your own model provider.
- LlamaIndex
The data framework for LLM apps, RAG-first, now with agent workflows and a document-parsing cloud.
The core framework is free open-source; LlamaCloud Free adds 10,000 parsing credits/month for managed extraction.
- OpenAI Agents SDK
OpenAI’s lightweight open-source framework for building agents, handoffs, guardrails and tracing in a few primitives.
Free and open-source (MIT), self-hosted — pay only for the model API you call.
- Pydantic AI
The type-safe agent framework from the Pydantic team, that FastAPI feeling for building AI agents.
Free and open-source, self-hosted — pay only your own model provider; the optional Logfire observability layer has its own free tier too.
- Mastra
The TypeScript-native agent framework — agents, durable workflows, memory and observability in one Apache-2.0 package, with an optional managed cloud platform.
Self-hosted (Apache-2.0) is free forever; the Starter Cloud tier adds 100K free observability events/month.
- Letta
A stateful-agent framework — open-source, memory-first agents whose context, tools and identity persist as a durable server object instead of resetting between sessions.
The Free tier runs a limited number of agents with your own API keys, ongoing.
- n8n
Source-available workflow automation with native AI-agent nodes, self-hostable and integration-rich.
The self-hosted Community edition is free forever under n8n's fair-code license.
- Relevance AI
Build and run an "AI workforce", teams of no-code agents that do real sales, ops and research work.
~200 free Actions/month with starter Vendor Credits — no card required to start building.
- Dify
Open-source, visual platform for building production LLM apps and agents, self-host it or run it in the cloud.
The self-hosted Community Edition is free forever (all core features); a Sandbox cloud tier adds 200 free message credits/month.
- Salesforce Agentforce
Salesforce's enterprise agentic-AI platform (rebranded "Agentforce 360") — prebuilt Service, SDR, Sales Coach and Commerce agents plus a low-code builder, running on live CRM data.
The Foundations tier includes 200,000 free Flex Credits plus 250,000 Data Cloud credits, ongoing.
- Perplexity
The AI answer engine, with agentic Deep Research and the Comet browser that acts on the web for you.
The Free plan includes limited Pro Search and a few Deep Research runs per day.
- Manus
An autonomous general AI agent that plans and executes multi-step tasks end-to-end in its own cloud computer.
~300 daily refresh credits free, no card required — roughly one task a day at zero cost.
- GPT Researcher
An open-source autonomous research agent that plans, searches many sources and writes a cited report.
Free and open-source (Apache-2.0), self-hosted with unlimited use — you pay only your own LLM/search API costs.
- Elicit
An AI research agent built for scientific literature, search, screen and extract cited data from 138M+ papers.
The Basic tier searches 138M+ papers and summarizes them unlimited, free, with limited Research Agent usage.
- Undermind
An AI literature-discovery agent that reads hundreds of full papers and follows citation trails, then reports how much of the literature it actually covered.
The Free tier includes AI chat, deep searches and reports with standard rate limits, ongoing — not a time-limited trial.
- Clay
GTM data platform with an AI research agent (Claygent), enrich, research and personalise outbound at scale.
500 free Actions/month plus 1.2K annual data credits, unlimited seats — the most generous free tier among sales agents here.
- Rox
An AI-native revenue operating system, "agent swarms" (one per account) that research, monitor and update your existing CRM across the whole deal lifecycle, with a genuine free self-serve tier.
The Starter tier includes 2,000 free agent actions/month, ongoing, for individuals running light pipeline generation.
- Unify
A single conversational GTM agent for warm outbound — self-serve from $0, built to augment a seller rather than replace one.
The Free tier supports up to 3 seats with limited credits across its full contact/company database.
- ElevenLabs Conversational AI
The voice-quality specialist’s own agent layer, 10,000+ voices and instant cloning, with a bring-your-own-LLM brain.
15 free call minutes/month across 4 concurrent calls, ongoing — LLM and telephony usage billed separately.
- LiveKit Agents
The open-source, self-hostable voice agent framework, Apache 2.0, provider-agnostic, and the infrastructure behind ChatGPT’s Advanced Voice Mode.
1,000 free agent-session minutes/month, 5 concurrent sessions, no card required.
- Zed
The open-source, GPU-native code editor built from scratch by Atom's creators — runs Claude Code, Codex and other agents via the open Agent Client Protocol it created. Free forever; Pro from $10/month.
The Personal tier is free forever — 2,000 accepted edit predictions/month plus unlimited use of AI features with your own API keys or external agents.
- Microsoft Agent Framework
Microsoft’s production-ready, open-source successor to AutoGen and Semantic Kernel — one agent SDK for .NET, Python and Go.
Free and open-source (self-hosted) across .NET, Python and Go — pay only your own model provider; the managed Foundry Hosted Agents layer is a separate, metered add-on.
- Google Agent Development Kit
Google's open-source, code-first agent SDK — four language runtimes, native A2A + MCP protocol support, and an optional managed Agent Runtime on Google Cloud.
Free and open-source (self-hosted, Apache-2.0) across all four language SDKs — pay only your own model provider; the managed Agent Runtime on Google Cloud layers on a metered add-on, not a requirement.
- Claude Agent SDK
Anthropic's own agent-building library — the same agent loop, tools and context management that power Claude Code, for Python and TypeScript. Claude-only, with no multi-provider escape hatch.
The SDK itself is free — self-hosted, pay only standard Claude API token rates for what the agent actually uses; the separate Claude Managed Agents hosted layer adds a metered $0.08/session-hour on top, not a requirement.