Which tools are on this page?
Tools whose price label is Free or Freemium, from a vendor page or an open-source license. Paid tools are excluded. If a pricing page could not be read, the tool is marked Pricing unverified and is not listed here.
Free and freemium
Best free AI tools
49 tools you can try without paying, grouped by what they do. If you're new to AI, start here. Tested means we tried it ourselves.
Coding agents
Editors, terminal agents, and prompt-to-app builders.
Tested
- S
Free coding helper. We asked it to write a small piece of code. It got it right in about 5 seconds and cost nothing.
- A
7/8 task runs passed, fast, $0. Small battery, one length-limit miss.
- C
Install and local capture work and are fast, but the core memory feature was not exercised. Provisional.
- C
4/4 completed tasks passed, but 4/8 were rate-limited. Provisional.
- F
Grade F. Cline did not write the code. Part of the failure was our test setup: we used a free anonymous server, not a normal paid model.
Not tried yet
- S
Shipped desktop code editor with an integrated coding agent. There is a free plan, and paid plans are listed on Cursor's pricing docs.
- S
Browser product built to generate a project from a prompt, with a freemium plan.
- A
In-browser app builder for generating and editing a project from prompts, with a public pricing page.
- A
GitHub's freemium coding assistant and agent, in the browser, editors, and a CLI.
- A
Vercel's freemium UI and app generation product. v0.dev redirects to v0.app.
- B
Cognition's software-engineering agent with a desktop editor and a cloud agent. The pricing page lists a $0 plan with a light agent quota.
Chat assistants
General chat and research assistants.
Tested
- A
8/8 on a small battery, fast, $0.
- B
Fast and correct on CPU. Compile step and shifting CLI flags add friction.
- B
6/8, misses instruction limits on length.
- B
Runtime installs fast and runs well on CPU. Scored on the runtime, not the small models' poor reasoning.
Not tried yet
- A
OpenAI's general chat assistant on the web and mobile, with a free tier. It is not a citation-first answer engine.
- A
Anthropic's chat assistant on the web and mobile, with a free tier.
- A
Google's chat assistant on the web and mobile, with a free tier.
- B
xAI's freemium chat assistant on the web and mobile. Filed as a general assistant, not a source-first research product.
Image
Still image generation.
Tested
- A
2 pass and 1 partial in 4 s each, free with no sign-up.
- A
Fast, clean cutouts on both images, with minor edge remnants.
- B
2 clean passes and 1 partial on a 3-prompt set, with correct text rendering. Small sample, slow free demo.
- B
3/3 perfect, but crashed on defaults and is slow on CPU.
- B
Excellent on documents. Failed scene text with defaults.
- C
Visibly sharper, but slow on CPU and less faithful by PSNR.
Not tried yet
- A
Freemium image product with generation plus Precise Edit, Remix, and Studio, documented by Ideogram.
- A
Freemium image product for concept art and general images, with a prompt-based editor.
Video
Text-to-video and image-to-video.
Tested
- C
Both clips coherent and fast, both missed prompt details.
Not tried yet
- S
Video generation and editing product. The pricing page lists a free grant of 125 one-time credits and 5 GB storage. runwayml.com redirects to runway.com.
Voice
Speech synthesis and cloning.
Tested
- A
Very fast and accurate on CPU. Fresh-install crash and clean-audio-only test cost points.
- A
Audio checked by transcribing it back. Not judged by ear.
- A
Accurate and fast with no dependency trouble. Synthetic audio only.
- B
Fast and intelligible on CPU. Naturalness not judged by ear.
Not tried yet
- S
Shipped voice synthesis, voice design, and dubbing product with a public pricing page and a free tier.
Agents and automation
Frameworks and workflow runners.
Tested
- S
Tier S on the 2026-10-20 CLI run: n8n 2.35.7 finished the HTTP-and-failure workflow, exit 0, wall clock 2.434s, cost $0.
- A
4/4 on a local, offline RAG check. Toy corpus.
- C
Install and local capture work and are fast, but the core memory feature was not exercised. Provisional.
- C
Happy path passed, failure case gave an empty reply, and local tool calling didn't work in our setup.
- D
Clean install and green tests, but the core claim is untested at $0. Provisional.
- F
Grade F. CrewAI did not finish the task. Part of the failure was our test setup: we used a free anonymous server, not a normal paid model.
- F
0/3 attempts passed. Install breakage and silent hangs.
Not tried yet
- A
Open-source platform for LLM apps and agent workflows. The cloud page lists a free tier of 200 message credits.
- C
Model API aimed at agent tasks, paid per token, with a free OpenRouter route that needs a key. It is not a workflow runner, and open weights are unreleased.
- D
Apache-2.0 inference framework for one DeepSeek model on NVIDIA SM100 or Huawei Ascend 950. It is not a workflow product, and the repo says KV cache, disaggregation, and load balancing are missing.
Gaming
Game assets, 3D, and playable games.
Tested
- B
Accurate shape fast, but heavy untextured mesh. 1 object only.
- B
6/6 tasks ready, fast and on-prompt. 3D splat quality not inspected.
- C
Works locally and follows prompt character. Slow on CPU, unheard.
- C
Fast, valid rigged GLB. Motion quality not viewed.
Not tried yet
- S
Production platform for image, video, 3D, and game-asset pipelines, with a public pricing page.
- A
Freemium text-to-3D and image-to-3D generator with a public pricing page.
- A
Web product whose job is turning a prompt into a playable game. The pricing page lists a free plan of 8,000 credits.
- B
Freemium concept-art and image editor. Public pages do not describe a sprite, texture, and 3D export pipeline.
FAQ
Why are some free tools marked Test pending?
Test pending means we haven't tried it yet. When we do, it moves into the Tested group.
Does free mean the model calls are free?
No. Open-source agents are labeled Free when the software is free. The note says when the model provider still bills you. A freemium label means the vendor page lists a free plan and a paid plan.