Free Quota Summary

Free version: The official website FAQ currently states "free daily generations" (free daily generation, not fixed at 25 points per month), but does not provide a specific daily limit; the partnership model (Nano Banana/Gemini 3, etc.) deducts points according to the original plan.

See full Adobe Firefly details →

Completely open-source and free (Apache 2.0), with no subscription model; uses local models (such as Ollama) at zero cost; API call fees depend on the chosen LLM provider

See full aider details →

10 credits (official website, as of 2026-08-22, retrieved from airmusic.ai/pricing, only lists credits, does not specify download or watermark details)

See full AirMusic details →

A free plan exists, but the official pricing page does not list its credit amount and the FAQ only spells out the restrictions: free users download and export content with watermarks (official wording), and outputs created under the Free plan are for non-commercial use only (official wording), so commercial rights are paid-only. Free credits come from promotions, and the official FAQ gives each type an expiry: sign-up credits expire after 31 days, daily check-in credits after 7 days and referral credits after 7 days, and once expired they are void. Credits included in a plan are issued at the start of each billing cycle, are valid for that cycle only and do not carry over if unused; top-up credits are sold to subscribers only and are valid for 365 days. The export resolution on the free plan is not published, while the lowest paid tier, Beginner, is 720P and 1080P starts at Creator. (Read from anijam.ai/pricing on 2026-08-09)

See full Anijam details →

25 message credits and 100 integration credits per month, with a maximum of 5 per day, on the US$0/month free plan. Message credits are spent when you give chat instructions that change the app, and integration credits are spent when the live app uses built-in services such as sending mail, generating images or calling an LLM. The official docs set a separate daily cap: you can use up to 5 credits a day, up to a total of 25 credits each month, so at most 5 message credits in one day. When the quota is exhausted, features that require credits are unavailable until your daily balance resets; when integration credits run out, any action you perform that requires integration credits fails and shows an error, and app users only see a generic error message. Unused quota does not carry forward: unused monthly credits expire at the end of your credit cycle and do not carry over to the next month. The free plan can create an unlimited number of apps, but has no custom domain and the result keeps Base44's default branding. (Verified 2026-08-09, source docs.base44.com/Account-and-billing/Credits)

See full Base44 details →

300,000 tokens per day and 1,000,000 tokens per month, US$0 (both figures appear on the official plan card and in the FAQ on the same page: Free plans have a 300,000 token daily limit and 1 million token monthly limit). Public and private projects can both be created; sites carry Bolt branding watermark text; single-file uploads are capped at 10MB; hosting is included, with up to roughly 333,000 web requests; the number of databases is unlimited. On resets, the official help centre states plainly that free-plan tokens reset on the 1st of each calendar month and do not roll over to the next month (only paid plans can roll over for one month). What happens once the quota is exhausted is not published; the site only says that upgrading to a paid plan removes the daily cap. (Read from bolt.new/pricing and support.bolt.new on 2026-08-09)

See full Bolt details →

Free version includes up to 200 Standard AI features or 20 Premium AI features, plus 5GB storage space and 1 brand kit (3 colors)

See full Canva AI details →

Basic editing is free, AI features require Pro

See full CapCut AI details →

NT$0/month, open to everyone. The official site lists the free-tier limits one by one: limited access to GPT-5.5 Instant, limited messages and uploads, limited and slower image generation, limited deep research, limited memory and context, limited Codex, limited ChatGPT Work desktop app. The only place the official site gives concrete numbers is the plan comparison table: free-tier GPT Instant total context window 27K (Go/Plus 54K, Pro 128K); free-tier single-input cap about 12 pages of text (Go/Plus about 40 pages, Pro about 250 pages); the context window for reasoning models is marked depends on the situation for the free tier (Go/Plus 256K, Pro 400K). Response time on the free tier is marked as limited by system resources and service conditions; only paid plans are fast. Chat history is unlimited on the free tier. Whether content is used for model training can be opted out of. The official site has never published a concrete messages-per-day figure, and this site does not fill in an estimate. (Read directly from chatgpt.com/zh-Hant/pricing over a Taiwan connection on 2026-08-09)

See full ChatGPT details →

The quota is calculated based on a rolling 5-hour usage window (not simply resetting at a fixed number every day), and general conversations can be used for around 10-20 times, excluding Claude Code.

See full Claude details →

US$0 per user per month with unlimited public and private repos, but PRs get only a summary rather than line-by-line review. Rate limits: 1 PR review per developer per hour, 3 in the IDE and 3 in the CLI, up to 150 files per review, and chat Q&A is not available (source: docs.coderabbit.ai/management/plans). There is also a separate free Open Source plan for open-source projects: 1 to 10 PR reviews per hour depending on how active the project is, 1 in the IDE, 3 in the CLI, 100 to 300 files per review and 25 chats. Newly registered organisations also get a 14-day Pro Plus trial with no credit card, dropping back to Free on the renewal date once it expires or is cancelled. The rate quota resets on a rolling hourly basis, and the site does not say whether you are queued once it is used up, only that a Fair Usage Policy applies. (Verified 2026-08-09)

See full CodeRabbit details →

A free ChatGPT account can use Codex, but with Limited access; the CLI version can be used with your own API key, and the fee is based on API usage.

See full Codex details →

The complete Chromium browser with basic AI functions (page summary, voice mode, shopping assistant, Deep Research) is completely free and has no usage limit (globally available starting October 2, 2025); however, agent tasks that consume quota (such as auto-form filling, booking, and emailing) have limited quota in the free plan.

See full Comet details →

Work email users: 500 API quota (30 requests/minute) + 10K Logo Link requests; Free email users: 250 quota (10 requests/minute). The quota is one-time and needs to be upgraded to a paid plan after it is used up.

See full Context.dev details →

10 credits per day, US$0/month, with the quota measured in credits. The same page also states that the free tier gets 5 collaborators and 3 collaborative Bots or Workflows. The cycle is daily, and the official site does not say whether unused credits accumulate. Per-call credit costs and free-tier usage counts (all from coze.com/premium): GPT-3.5, Claude 3 Haiku and GPT-4o mini are all 0.1 credits per call, about 100 calls per day; Gemini 1.5 Flash 0.25 credits per call, about 40 per day; GPT-4o(8k), Claude 3 Sonnet, Claude 3 Opus and Claude 3.5 sonnet_v2 are all 2 credits per call, about 5 per day; Gemini 1.5 Pro 2.5 credits per call, about 4 per day; DeepSeek-R1 3 credits per call, about 3 per day; GPT-4 Turbo, GPT-4o(128k) and GPT-5 are all 5 credits per call, about 2 per day. Plugin costs: Bing Web Search 1, Stable Diffusion 1, WebPilot 2, Leonardo 2.5, DALLE 3 10 credits per call. Note for users in Taiwan: the top of the official page states Payment service currently unavailable in Russia and Taiwan, so paid upgrades are currently unavailable in Taiwan and the Premium Lite button simply shows Payment not supported. (Read from coze.com/premium on 2026-08-09)

See full Coze details →

The official website does not disclose the free quota; charges are based on actual usage (no minimum consumption, no monthly fee)

See full CreateOS Sandbox details →

Hobby: 2000 times of completion + 50 times of slow progress requests (confirmed on the official website on 2026-07-26, but the official website no longer lists these numbers in detail, only mentioning limited quota, and the specific numbers are now only seen in third-party community sources, keeping the record but adjusting the confidence level to medium)

See full Cursor details →

50,000 characters and 1 file translation per month (file size limit: 5MB). Checked on 2026-08-22, the previously recorded dynamic loading issue has been resolved, as confirmed by a first-hand check on the official website (deepl.com/en/pro) in the Taiwan (USD $) region.

See full DeepL details →

The web version is basically free, but has peak traffic limits.

See full DeepSeek details →

60 media minutes, that is 1 hour, per month plus 100 one-time AI credits. Media minutes means the length of material imported or recorded and handed to Descript to process. The biggest hurdle on the free plan is not the minutes but the AI credits: there are only 100 and they are one-time rather than granted monthly (the free column of the official comparison table reads AI credits 100 (one-time)), so once they are gone they are gone unless you upgrade. All AI tools, including Underlord, Studio Sound, green screen, Eye Contact, filler-word removal, image and video generation and voice cloning, are marked Limited on the free plan precisely because of this 100-credit cap. Export limits: local export and web share links are both 720p, web video exports are capped at 1 hour, and audio exports have no length limit. Cloud storage 5GB and 1 seat; projects and animated captions are unlimited. Transcription supports 25 languages and can identify 8 or more speakers. No credit card is required. Watermark: the official comparison table has a Watermark-free video export row but the free column shows only an icon with no wording, while the paid copy says 1080p, watermark-free; three third-party sources agree that free exports are 720p with a watermark, so this point is flagged as third-party evidence only, not stated in words on the official site. The reset date for the 60 minutes and commercial licensing on the free plan are not published. (Read from descript.com/pricing on 2026-08-09)

See full Descript details →

1,000 credits per month (not simply character count, shared among various features such as TTS/voice cloning, equivalent to approximately 10,000 characters for TTS, with licensing restrictions)

See full ElevenLabs details →

The free tier offers unlimited recording, unlimited transcription, and unlimited storage with no expiration date. Advanced AI summary is available for the first 5 meetings per month, and then downgraded to a basic summary template afterwards (as of 2026-08-28, according to the official pricing page, the 5-meeting limit for summary is from a secondary cross-source).

See full Fathom details →

The free version has unlimited basic search, and approximately 5 Pro searches per day (200 points per day, resets at midnight), as well as a daily quota of 3 file analyses (confirmed on the official website on 2026-07-30); the paid Pro plan, after using up 15,000 points per period, can still enjoy up to 300 Pro searches per day (confirmed on 2026-07-26: the latest blog post on the official website states 300 times, while the FAQ page states 100 times, which is outdated and not updated).

See full Felo details →

On the official pricing page for the free tier, the Free column reads: “Unlimited transcription*”, “Limited AI summaries”, “400 mins of storage/team”, “20 AI credits” (2026-09-04 direct read fireflies.ai/pricing). ⚠️ The AI summaries are Limited, not unlimited; the original “Unlimited AI summaries” was a typo, corrected here; Unlimited AI summaries are only available from the Pro plan. The reset cycle for the 20 AI credits: the official free‑plan description clearly states “20 AI credits per month”, meaning a monthly reset (correcting the earlier note that “the website did not disclose the reset cycle”). The 400‑minute storage is a total amount shared per team, not a monthly reset; when full, users must delete older meetings to store new ones. The free‑plan description also mentions “per seat”, which conflicts with the pricing page and the official blog’s “per team”; the pricing page is taken as authoritative. Transcript download (Download transcripts) is listed as a Pro‑only feature; the free version cannot export transcripts; single file uploads are limited to 100 MB/150 minutes.

See full Fireflies.ai details →

The free tier offers 3 credits per month (approximately 5 minutes of audio/video), with output at 720p and a Fliki watermark, and no commercial use license (Checked on the official pricing page on 2026-08-28; a secondary source claims the free quota is 36 credits per month with a 1-minute limit per clip, but this is inconsistent with the official page, which is taken as authoritative and noted here)

See full Fliki details →

100 actions per day using your own AI, on the US$0/month free plan. The build quota is 100 actions per day using your own AI such as Claude or ChatGPT, where an action is one edit, read or test performed by the AI assistant, and the site notes that failed actions never count. The quota resets daily, and the free plan also has a weekly pool (official wording: Limits reset daily (Free also has a weekly pool)). In-app AI features get a further 3,000 starter credits, and hosting gets a US$3 starter allowance to get your first app live. The official FAQ adds further limits: a free account keeps at most 2 projects and 1 published app, and can only use the free *.floot.app subdomain, while custom domains and removing the Floot badge both require an upgrade to Pro. Rollover credits require a paid plan. (Read from floot.com/en/pricing and the official docs on 2026-08-09)

See full Floot details →

Free version: unlimited design drafts, publish to framer.website subdomain (with Framer badge), 1 GB bandwidth, 500 AI credits per month, 5 MB file upload, 1 language, up to 3 editors; custom domain requires paid site plan (Checked on official pricing page as of 2026-08-28)

See full Framer details →

Register to receive one-time points

See full Gamma details →

NT$0/month, free to anyone with a Google account. Model available: Gemini 3.6 Flash; the official site states plainly that access to 3.1 Pro may vary, meaning free-tier access to Pro fluctuates and is not guaranteed. Features included on the free tier: image generation and editing, Deep Research, Gemini Live, Canvas, Gems, Gemini Notebook (research and writing), the Google Flow creative studio, plus limited usage of Nano Banana Pro. Cloud storage 15 GB (shared across Gmail/Drive/Photos). How the quota is counted and when it resets, in the official footnote's own words: the Gemini app measures usage limits by compute, which depends on prompt complexity, the features you use and conversation length; usage resets every 5 hours, up to a weekly usage cap. In other words the free tier is not a fixed number of messages per day but a compute allowance that resets every 5 hours plus an overall weekly cap; long conversations, Deep Research and image generation burn through it faster than plain chat. When the quota runs out you can buy AI credits. The official site does not publish the specific compute figure for the free tier. (Read directly from gemini.google/subscriptions over a Taiwan connection on 2026-08-09)

See full Gemini details →

The free version is sufficient and permanently free, with limits on the number of sources or notebooks, and does not require subscription or credit card.

See full Gemini Notebook details →

The general user free tier includes 2,000 completions and 50 chats per month. Open-source maintainers and teachers can upgrade to Pro for free. The free student plan has reopened for new applications since 2026-06-17, after being suspended since 2026-04-20 for 58 days. As of 2026-07-27, the model selection for Free and Student plans has been limited to only Auto, and manual model specification is no longer available, while Pro and above plans are not affected.

See full GitHub Copilot details →

120 credits once at sign-up, with no recurring free allowance: the free tier is a one-time launch welcome pack rather than a monthly quota. The official FAQ puts it bluntly, saying there is no recurring allowance on Free and that building stops once the balance reaches zero, and that this welcome pack is a one-time allocation which does not roll over or recur. Credits are deducted by actual workload, and the site says the more work a turn involves the more it costs, so small edits are cheap and building an app from scratch is dearer; it is not a flat per-message charge. The free plan can browse, install and run apps from the public store, which the site says is entirely free, and can also publish your own apps to the public store. The free plan cannot publish as unlisted, cannot prevent others from remixing, and cannot buy extra credits. Platform limit: currently macOS Tahoe with Apple Silicon only. There is also a 14-day Store Trial for testing Team's private team store. (Read from glaze.app/pricing on 2026-08-09)

See full Glaze by Raycast details →

Development purposes are almost free with rate limits, and the exact RPM/TPM/RPD depends on the account level as stated on aistudio.google.com/rate-limit. For the Gemini 2.5 series, Search grounding is free for 1,500 times/day, and then US$35/1,000 times, while Maps grounding is also free for 1,500 times/day, but then US$25/1,000 times (confirmed on the official website on 2026-07-27 that the excess fee rates for the two are different). Other Search/Maps models are free for 500 times/day. For the Gemini 3 series models, the first 5,000 prompts per month are free (both Search and Maps), and then US$14/1,000 times for queries.

See full Google AI Studio details →

Grammar and spell check, tone detection, 100 times of AI generation per month (Checked on grammarly.com/plans as of 2026-07-31)

See full Grammarly details →

The official site publishes no concrete usage figures at all for the free tier, and this site does not fill in an estimate. Verification trail (exhausted): grok.com/pricing returns HTTP 200 but the whole page is rendered by JavaScript, so the static HTML contains no plan names or price text, only Stripe price IDs; opening it in a real browser hits a Cloudflare human-verification wall, which we do not bypass; x.ai/grok returned 403 three times in a row (blocked, not 404). What can be confirmed: free use does exist and is US$0, because in the in-app purchase list on the official Apple App Store listing every paid item belongs to the SuperGrok range, so basic chat requires no payment. Third-party felloai.com states plainly that xAI has never published free-tier numbers anywhere, and says the free tier only gets the lighter Grok 4 Mini, includes basic real-time web and X search, has limited voice mode, and that image and video generation were removed from the free tier on 2026-03-19. The widely circulated figure of about 10 messages every 2 hours has no official backing and is not used here. (Verified 2026-08-09)

See full Grok details →

The Free Plan needs no card and each model has its own per-minute and per-day limits: Llama 3.1 8B Instant is 30 requests per minute and 14,400 per day, 6,000 tokens per minute and 500,000 tokens per day; Llama 3.3 70B Versatile is 30 per minute and 1,000 per day, 12,000 tokens per minute and 100,000 tokens per day; GPT OSS 120B and 20B are 30 per minute and 1,000 per day, 8,000 tokens per minute and 200,000 tokens per day; Groq Compound is 30 per minute and 250 per day, 70,000 tokens per minute; Whisper Large V3 is 20 per minute and 2,000 per day, 7,200 audio seconds per hour and 28,800 audio seconds per day. Going over the limit returns HTTP 429, and cached tokens do not count towards the limits. (Read directly from the official docs at console.groq.com/docs/rate-limits on 2026-08-09)

See full Groq details →

官網未公布具體數字：定價頁僅標示「Free Plan, No Credit Card」，FAQ 亦只提到取消訂閱後仍保有「有限的浮水印生成」，未列免費額度數字

See full Hedra details →

3 videos per month, each capped at 1 minute, on the US$0/month free plan; the quota is counted in videos, not credits (the official FAQ confirms again: The Free plan includes up to 3 videos per month, no credit card required). Quality up to 1080P (4K needs Pro or above), processing at the slowest Standard speed. Assets available: 500+ built-in avatars, 1,000+ AI voices, 75+ video templates, 1 custom avatar, 1 voice clone, up to 3 photo avatars, 30+ languages (paid plans get 175). Avatar IV and Video Agent are accessible but still bound by the 1-minute limit; video translation is likewise capped at 1 minute per video. 1 seat. The official feature table lists remove watermark from Creator upwards and leaves that row unticked for the free plan, so free-plan output carries a HeyGen watermark. The free plan has no monthly credit allowance and is not eligible for credit rollover. What happens after the 3 videos are used, and commercial rights on the free plan, are not explained on the official site. The quota resets monthly. (Read from heygen.com/pricing including the feature table and FAQ on 2026-08-09)

See full HeyGen details →

The open‑source self‑hosted version uses a Modified Apache 2.0 license and BYOK (bring your own key); the official site states “Free to run, yours to change”. The desktop version’s website says “free to start”, but it does not disclose specific free usage limits (neither number of runs nor credits are listed). Paid side: the Growth plan offers a 7‑day free trial (the site says “Growth is free for 7 days”).

See full holaOS details →

The official website ideogram.ai/pricing confirms that it is 10 slow credits/week, which means 10 slow credits per week, not per day, and can generate approximately 40 images/week.

See full Ideogram details →

The free tier has approximately 10 minutes of AI generation per week, exports are watermarked, and the number of exports per month is limited (around 4 times), with a quota reset on Mondays; the official pricing page no longer publicly lists the details of the free tier, only listing paid plans (Checked on 2026-08-28 from the official pricing page)

See full InVideo AI details →

Daily login rewards points, can generate a small amount

See full Kling details →

Free desktop version includes 60 minutes of noise reduction per day and 2 AI meeting notes per day, with unlimited transcription, according to secondary sources, but the official pricing page only shows 7-day free trial with full features (Checked on 2026-08-28 by reading the official pricing page)

See full Krisp details →

150 Fast Tokens per day, expiring in 24 hours without accumulation or carryover, and only 1 generation task can be executed at a time

See full Leonardo AI details →

Free version is US$0 per month, with a limit of 3 notebooks (not unlimited), for single-user use, 4GB RAM, 0.5 vCPU, and a US$5 AI quota limit (cannot be topped up once used up)

See full Livedocs details →

5 build credits are issued per day, capped at 30 per month, plus 20 Cloud credits per month and 4 credits per month usable for in-app AI features (both the pricing-page FAQ and the official docs at docs.lovable.dev state this, giving two cross-checked sources). The point most easily misread: credits are not a number of chats. In Default mode credits are deducted by task complexity, and the official examples are 0.50 credits to make a button grey, 0.90 to remove the footer, 1.20 to add login authentication and 1.70 to build a landing page with images; Plan mode is a flat 1 credit per message. So how many instructions 5 daily credits buys varies. Credits expire: the free plan's daily allowance disappears if it is not used that day and does not accumulate, and the 30-per-month cap is hard. The free plan includes Git sync, an unlimited number of workspace members, projects private to the workspace, and community support. Code and AI output are owned by the user (official FAQ: You do. You own your code), and the site sets no commercial restriction on the free plan. (Verified 2026-08-09)

See full Lovable details →

⚠️2026-08-22 Two checks: The official website's pricing page no longer has a separate Free plan card (only Creator/Pro/Business/Custom four paid plans), only the phrase "Sign Up - It's Free" is seen at the top, unable to confirm whether the previously recorded "10 daily Image generation/3 daily Video generation free quota" still exists, the indication is uncertain, waiting for the next check to clarify through other channels (in-app or FAQ)

See full Magic Hour details →

Free tier has 1,000 credits per month, up to 2 active scenarios, minimum scheduling interval of 15 minutes, maximum execution time of 5 minutes, and file size limit of 5 MB (Checked on 2026-08-28 from official pricing page)

See full Make details →

300 credits per day (reset every 24 hours), supports Chat mode and Agent mode (Manus 1.6 Lite), 1 task at a time; new users also receive 1,000 one-time starting credits

See full Manus details →

The web version offers basic text conversation for free; advanced features such as Copilot Chat, Researcher/Analyst agents, and image generation in Office require Microsoft 365 Premium (US$19.99/month) or an M365 subscription.

See full Microsoft Copilot details →

About 25 messages per day: Mistral Vibe's free tier has a limited daily message quota, and includes web search and code features

See full Mistral AI details →

Completely free, with no pricing page, no paid plan and no published usage cap. There is no pricing page, paid plan or subscription information anywhere on the official site; joining and using it are free, with no credit system and no published usage limit. How to join: first let your own AI agent read the instructions at moltbook.com/skill.md, the agent registers itself and produces a claim link, and you then verify ownership by posting from your X (Twitter) account, which completes the binding. Once bound, the agent connects automatically through the site's Heartbeat mechanism to browse, post and comment, with one heartbeat every 4 hours. Human users are positioned as observers, free to browse content but not to post. The site also has a Build for Agents section offering developer features for agent verification using a Moltbook identity, on an early-access application basis with no published pricing. Because there is no paid plan, there are no questions of watermarks, commercial restrictions or quota resets. (Read from moltbook.com on 2026-08-09)

See full Moltbook details →

US$0: 40 standard-model uses per day on a shared quota across Claude 4.5 Haiku / Gemini 3.1 Flash-Lite / GPT-5.4 mini, limited trial of image and video generation, limited trial of ChatPDF, time-limited AI Agent trial, up to 3 logged-in devices (read directly in a browser from the Traditional Chinese monica.im/pricing on 2026-08-09)

See full Monica details →

10 projects, 10 minutes of generation quota, no download rights

See full Murf AI details →

(2026-08-28 checked on the official pricing page) The free version has two options: ① self-hosted Community Edition = permanently free (source code on GitHub, Sustainable Use License, available for internal commercial use, no execution limit, but excludes enterprise features like SSO/environment management); ② official Cloud has no permanent free tier, only a 14-day free trial (trial of Pro plan with full features, limited to 1,000 executions, no credit card required), and workspaces will be deleted if not upgraded after expiration (workflows can be downloaded as a backup within 90 days).

See full n8n details →

The free version of Gemini includes a small amount of Nano Banana generation quota; paid plans increase the quota limit based on the level (official website confirmed on 2026-07-30 that the official name is "Nano Banana Pro", with a tiered system of Plus = limited access, Pro = more access, and Ultra = higher or highest access, but the official website does not disclose the specific quota limit, only estimated by third-party sources, not officially confirmed).

See full Nano Banana details →

Free：每週 500 點 AI credits（官網寫 500 AI credits per week，週週重置）；含無限視覺編輯、無限檔案匯入（PPT／DOC／PDF／HTML／MD）與無限 PNG、PDF 匯出，但帶 Napkin 浮水印品牌標記

See full Napkin AI details →

The free version has limited AI functionality (trial quota for Q&A), and full AI functionality (Notion Agent, AI meeting recording, etc.) has been integrated into the paid plan. The free version also includes up to 10 external guests and 7-day page history.

See full Notion AI details →

More than 25 free models, 4 free providers, daily limit of 50 requests, community support (Checked on openrouter.ai/pricing as of 2026-07-31)

See full OpenRouter details →

Self-hosted version is completely free (requires self-paid DataForSEO API fee; the official website does not disclose the starting price, and the original note on the website of about US$6/month cannot be found on the official website, pending verification); the hosted version has a trial quota of US$0.50.

See full OpenSEO details →

The free version (listed as Free forever, US$0 on the official website) has a quota of 60 points per month, with the official comparison table stating that 1 point corresponds to 1 minute of source video processing. It can output videos up to 1080p, automatically reframe, generate AI subtitles (including emojis and keyword highlighting), support multilingual transcription, 9:16 vertical format, and can be directly posted to social media. However, there are three major limitations: 1. Has watermark - all outputs from the free version have the OpusClip watermark; 2. No editing - it does not provide an editor, and thus cannot use timeline or text editing, change font, add B-Roll, or schedule posts; 3. The materials and finished products are only saved for 3 days (as stated on the official website: After 3 days, the clips will no longer be exportable; the comparison table's Media storage column also states Expires after 3 days), and cannot be exported after that. The processing speed of the free version is general (paid versions have priority); it can only upload local videos (with a single file limit of 10GB) and cannot automatically import from YouTube. It is not stated on the official website whether the 60 points quota resets daily or if unused points accumulate; commercial use authorization for the free version is also not mentioned. (Checked on 2026-08-09 from opus.pro/pricing)

See full Opus Clip details →

免費版每月 60 credits、免信用卡：輸出最高 1080p、自動重構圖、AI 字幕（含表情符號與關鍵字強調），但成品帶浮水印、不能進編輯器、剪出的片段 3 天後就不能再匯出（2026-09-05 本機瀏覽器直讀 opus.pro/pricing）

See full OpusClip details →

300 minutes of transcription quota per month, single meeting recording up to 30 minutes, 3 lifetime audio/video file imports (Checked 2026-07-31 from official website otter.ai/pricing)

See full Otter.ai details →

Daily basic searches are unlimited, Pro searches are limited to a small amount per day

See full Perplexity details →

80 video credits per month, US$0; the free plan is called Basic and credits reset monthly. Three explicit caps: (1) resolution only up to 480p (official wording: Access to Pika 2.5 (480p only)), while 720p and 1080p require Standard or above; (2) image-to-video only, so text-to-video is not available; (3) rollover credit top-ups are not allowed. What the free plan does include: Pikascenes, Pikadditions, Pikaswaps, Pikatwists, Pikaffects, plus the two items listed on the plan card itself, watermark-free downloads and commercial use. How far 80 credits goes, from the official credit-cost table for free files: text-to-video and image-to-video 480p 5 seconds 12 credits, Pikascenes 480p 5 seconds 20 credits, Pikaffects 5 seconds 15 credits, Pikatwists 5 seconds 60 credits, which works out at about 6 five-second 480p clips, or a single Pikatwists run that almost empties the balance. The official site does not say what happens once the quota is used up. Note: third-party eesel.ai claims the free tier has watermarks and no commercial use, which directly contradicts the official plan card; the official site is taken as authoritative. (Read from pika.art/pricing on 2026-08-09)

See full Pika details →

Second-hand information disclosure: the free-quota figures below are not taken from the official pricing page. The official pricing page sits behind a login at app.pixverse.ai, and four public paths all returned 404 in testing, so no first-hand figure could be obtained. The official blog at pixverse.ai/en/blog/best-free-ai-video-generator only confirms an entry path that gives registered users free credits, explicitly says the amount can vary by account, region, and product route, and tells users to check their balance in the app; the same post states the V6 rates as 9 credits per second for 720p without audio and 12 credits per second for 720p with audio. Several third parties, namely atlascloud.ai, somake.ai, tooljunction.io and costbench.com, consistently record a one-time 90 credits on sign-up plus 60 credits per day, resetting at 00:00 UTC with unused credits void that day, free output watermarked and resolution-limited, personal non-commercial use only, commercial use requiring Standard or above, and credits not carrying over between months. This site is only passing the information on, it does not represent this site's position, and the official announcement prevails. (Verified 2026-08-09)

See full PixVerse details →

2026-07-26 update: The daily free points have been reduced from 3,000 points to approximately 300 points (non-cumulative), which took effect around late March 2026. This information is confirmed by two independent secondary sources (aidetectplus.com and carletontorpin.com), but the official pricing page cannot be directly verified due to repeated WebFetch failures, only obtaining cookies and navigation pages. Therefore, this figure is confirmed by secondary sources, not directly verified by the official website.

See full Poe details →

(2026-08-28 直讀官方定價頁) 免費層：改寫每次最多 125 字、僅 Standard/Fluency 2 種模式；文法檢查僅基本錯誤；Humanizer 每次 125 字、每日 6 次；摘要上限 1,200 字（僅基本摘要）；AI Detector 受限存取；抄襲檢測不含（付費 25,000 字/月）；AI Chat 20 次對話；AI 生圖 3 張/日；文件儲存 10 MB；無同義詞滑桿調整、無自訂模式。官方註明各平台限制可能不同。

See full QuillBot details →

Free version US$0 (checked official pricing page on 2026-08-15): 3 Image generations per style per day, 3 image uploads per day, up to 2 images at a time; limited to personal use, no commercial use; all generated images are public; only Recraft's own models can be used, no parallel generation

See full Recraft details →

Daily free Agent credits, amount unpublished; the free plan is called Starter. The official site only says daily free Agent credits, and how many credits, the daily cap and the monthly total are published neither on the pricing page nor in the official docs (the docs only say Starter users get daily Agent usage credits, with a monthly cap, reset daily). The concrete limits that can be confirmed: (1) Agent can only use Lite build, while Full build needs Core or above; (2) only 1 project can be published, and that published link goes offline automatically after 30 days; (3) published apps carry a mandatory Made with Replit badge, removable only on Core; (4) 2GB workspace storage; (5) there is a separate monthly cloud credit allowance for databases, object storage and deployments, likewise unpublished, reset monthly; (6) no third-party connectors, no Replit AI Integrations, no Plan Mode. The free plan can use the built-in database, can make slides, videos and animations, and can publish private or password-protected deployments. (Read from replit.com/pricing and docs.replit.com on 2026-08-09)

See full Replit details →

Free:每月 10 次 AI 問答(10 AI answers per month)、單一儲存庫上限 250 MB、commit 分析上限 500 筆、可索引 2 個儲存庫,且僅限公開儲存庫(官網敘述為「Built for open source」),含每週自動更新徽章

See full Repowise details →

A one-time 125 credits, US$0/month; the plan is called Free and the official site labels it Free forever. The core quota is a one-time 125 credits, not a monthly grant; the official FAQ says these 125 credits never expire, but they also never reset, so once they are gone the only way to generate more is to upgrade or buy credits (top-ups start at 1,000 credits). Asset storage 5GB. Only some generative models are included for testing concepts, not the full model line-up. What 125 credits actually buys, using the site's own conversion figures: Gen-4.5 costs 60 credits per 5 seconds, Gen-4 Image 1080p 8 credits per image, Turbo 2 credits per image, which is roughly 2 five-second Gen-4.5 clips, or 15 images. The official site lists No watermarks and 4K upscaling from Standard upwards, and those two rows are not listed for the free plan. Commercial-use rights for the free plan are not explained on the pricing page. (Read from runway.com/pricing including the on-page FAQ on 2026-08-09)

See full Runway details →

10,000 characters of AI content generation per month, Free forever and no credit card needed. The unit is characters, not words and not runs: 10,000 characters is roughly 3,000 to 5,000 Chinese characters or about 1,500 to 1,800 English words, so it runs out quickly. Free-plan limits, column by column from the official comparison table: only 1 language, against 35 or more on Premium; the Tone of Voice Match column reads No Tone Match; the plagiarism-check column reads None, meaning none at all; and the character input limit column is a dash, meaning no doubling. The free plan still gets the Chrome extension, 20 or more built-in tones, the dedicated community and custom use cases. Ownership: the official FAQ states plainly that output belongs to the user or to the company using Rytr, without distinguishing free from paid, so free-plan output is yours too, subject to the terms of service. Whether the quota resets on the calendar month or on the sign-up date, what you see when it runs out, and whether you can top up, are all unpublished. (Read from rytr.me/pricing on 2026-08-09)

See full Rytr details →

CLI 本身開源（Apache-2.0）免費；SandBase 平台提供起步免費點數（Start：Free credits，官網定價頁未公布點數數額，本站不編），之後依所呼叫的模型／API／Agent 實際用量計費（2026-09-05 本機瀏覽器直讀 sandbase.ai/pricing）

See full SandBase CLI details →

Free plan US$0, "Free forever". Included credits are "1,000 one-time", i.e. a single 1,000-credit grant (not monthly); the Weekly refresh column shows "–", meaning no weekly top-up (only paid plans get Pro +2,000/week and Max +4,000/week, reset every 7 days and not accumulated). 1 workspace, cannot invite teammates. Concurrent executions capped at 10; rate limits of 50 sync runs/min, 200 async runs/min and 30 API endpoint calls/min; execution timeouts of 5 minutes (sync) and 90 minutes (async); storage of 5GB of files, up to 5 tables, up to 50,000 rows per table, 30 days log retention. Not included on the free plan: Sim Mailer (Inbox), KB Live Sync, Sandboxes, Slack Connect, Access Control, SSO, SOC2 Compliance, Self Hosting, Dedicated Support. When credits run out, the official docs state "usage is capped at your plan's included credits and runs will stop" (runs stop by default; paid plans can enable on-demand overage at US$0.005/credit). 1 credit = US$0.005 (docs). The source conflict recorded on 2026-08-09 ("pricing page monthly vs docs one-time") is resolved: on 2026-09-16 both the pricing page and docs.sim.ai/platform/costs show 1,000 one-time / Pro 6,000 per month / Max 25,000 per month. Sim is an open-source project. (Read directly from sim.ai/pricing and docs via playwright on 2026-09-16)

See full Sim details →

Free Plan US$0 'Forever Free': 1 team member, 1 social account, 30 days of analytics retention, 5 competitor accounts monitored per platform, scheduled posting, manual replies to comments and DMs, and AI creation features available. There is also a 7-day free trial (the official FAQ states no credit card is required)

See full SocialEcho details →

The free version can be used long-term but has clear limitations: only 10 basic (mechanical) voices, a reading speed limit of 1.5 times, and a file storage limit of 5 files; 1000+ natural human voices, 5x speed, OCR scan reading, and offline download are all locked in Premium (Checked official pricing page on 2026-08-28)

See full Speechify details →

Open-source model weights are available for free self-hosting (Stability AI Community License, commercial use is allowed if annual revenue is less than US$1 million, and a graphics card is required); the official website also offers a Brand Studio service with 1,000 free trial points.

See full Stable Diffusion details →

50 credits reset every day on the US$0/month free plan, which the pricing page calls Free Plan (the help centre still calls it Basic). The quota is daily, not monthly. A footnote at the bottom of the official page specifically states that credits included with a subscription do not roll over daily or monthly, so any of the 50 credits left unused that day are gone the next. Only the v4.5-all model is available (the official site calls it the best free model); v4, v4.5, v4.5+, v5 and v5.5 are paid-only. 4 songs generate in parallel in a shared queue (paid gets 10 songs with priority queueing). Audio uploads are capped at 8 minutes (30 minutes on paid). On features the free plan is explicitly listed as standard features only, no stem separation and no credit top-ups, so once the 50 daily credits are gone you either wait for tomorrow or upgrade. Most important is licensing: the official site says No commercial use, and the help centre goes further, stating that songs made on the free plan are owned by Suno and users are only permitted to use them non-commercially. (Read from suno.com/pricing and help.suno.com on 2026-08-09)

See full Suno details →

Monthly 10 minutes of video, 9 AI virtual avatars, 1,200 credits/month, 25 AI-generated video assets, 1 editor, finished products have watermark; Personal Avatars free version not available (official website comparison table only has 3 values: 3/5/Unlimited, corresponding to Starter/Creator/Enterprise)

See full Synthesia details →

Free tier includes 1 user, up to 3 AI apps, one-time 6,000 AI credits (1,000 for email verification, 5,000 for initial setup, non-renewable and non-refillable), 1 AI agent, 3 AI automations, 250MB storage space, and 500+ community templates (Checked on the official pricing page as of 2026-08-28)

See full Taskade details →

200 credits, 100 automation runs per month and 1,000 rows per space, at US$0 per seat per month, For individuals. Credits are the unit for AI features, covering AI Field, AI Chat, CuppyClaw and App Builder; each automation run uses 1 of the 100 monthly runs; the 1,000-row cap is per space, while rows in a single table are unlimited; attachment storage is 1GB per space. Features included on the free plan: AI chat, AI field, App builder, CuppyClaw (full create, read, update and delete on your data by chatting in Slack, Telegram or WhatsApp), all of the Grid, Form, Kanban, Gallery and Calendar view types, all basic field types, Dashboard, Trash restore, templates, CSV and Excel import and CSV export, and the Web API (rate-limited to 10 requests per second, the same as paid). Free-plan restrictions: automation run history is kept for only 2 weeks and row change history for only 2 weeks (paid keeps 1 year or 3 years); no Authority matrix, SSO, domain verification or custom app domain. The reset cycle for credits and what happens once they run out are not clearly published. (Read from teable.ai/pricing and teable.io/pricing on 2026-08-09)

See full Teable details →

Free version (as of 2026-08-15 from official pricing page): 5,000 times of auto-completion per month, limited basic usage, standard queue, TraeWork 2 concurrent cloud tasks; available for download and use

See full TRAE details →

The free version offers 10 points per day + 100 points per month (non-cumulative), which can generate approximately 1-3 songs per day, each within 2 minutes.

See full Udio details →

The free plan is US$0/month, with a quota that is a dual limit of "points + daily messages", and stops when either limit is reached: ① it includes a US$5 point balance per month (the official documentation also states Monthly Credits = US$5); ② a daily limit of 7 messages (the official website states 7 message/day limit). The points are consumed as follows: v0 is billed based on tokens, and the official website lists the prices for each model - v0 Mini input costs US$1/1 million tokens, output costs US$5; v0 Pro costs US$3/US$15; v0 Max costs US$5/US$25; v0 Max Fast costs US$10/US$50. This means the US$5 quota on v0 Max can only generate about 200,000 output tokens. The behavior when points are used up is clearly stated in the official documentation: generation will be paused, and points cannot be purchased for the free version (purchasing is only available for Premium/Plus/Business). Points will be rolled over to the next billing cycle, but will expire after 65 days. The free version can still be used for: deployment to Vercel, Design Mode visual editing, GitHub synchronization, and basic usage analysis. (Checked on 2026-08-09 from v0.app/pricing and v0.app/docs/pricing)

See full v0 details →

As of 2026-09-16 the official pricing page has been restructured into "Usage only (No Success Package) + optional Success Package". The sign-up starting allowance is "US$5 free credits to get you started" (US$5 of prepaid credit, not minutes; credits are prepaid US dollars applied to Vapi hosting at US$0.05/minute plus model, voice, transcription and telephony costs, the latter four passed through at provider list prices with no markup). Usage only includes 4 concurrent calls, 2 organizations, 1 included Vapi phone number, 14 days raw data retention, full model catalog, full platform access, peer-to-peer support via the Discord community, and email support without SLA. When the balance reaches 0 and auto-reload is off, the subscription is frozen and no new calls can be started (docs.vapi.ai billing). The old wording "60+ minutes included", "10 concurrent lines included" and "US$10 credit" no longer appears on the official site. (Read directly from vapi.ai/pricing and docs.vapi.ai via playwright on 2026-09-16)

See full Vapi details →

US$0 with a VEED watermark on every export, a 10-minute export length cap and 720p quality cap. Hard limits item by item, from the Free column of the official See all features comparison table: every exported video carries a VEED watermark (the FAQ on the same page says the watermark appears in the corner of videos made on the free plan and that removing it requires any paid plan); single-video export length is capped at 10 minutes; export quality is capped at 720p, with 1080p and 4K paid only; cloud storage 2GB; single-file upload cap 1GB; the stock library is Limited. Free AI quotas: auto-subtitles 10 minutes per month; text-to-speech up to 10 minutes per month, using credits; Gen-AI Studio video generation up to 12 seconds per month; AI image generation up to 100 images per month, using credits. AI editing features such as Voice Clone, Text Behind Person, AI B-Roll, Clean Audio, filler-word removal, silence removal, Eye Contact and background removal are all marked Try once on the free plan. Resets: the official FAQ says credits refresh in full at subscription renewal and do not roll over, and subtitle and translation minutes likewise do not carry over. Commercial licensing on the free plan is not published. (Read from veed.io/pricing on 2026-08-09)

See full VEED details →

15 Velo minutes per month, with 1 seat, 1 voice clone and 1 avatar video per month. Minutes are consumed by generation, rendering, voiceover, translation, avatars and agent workflows. On features the free plan gets the full editor, prompt-to-video, document-to-video, screen-recording clean-up and basic support. Watermarks work two ways: sharing by link is watermark-free (official wording: Share videos without watermarks), but downloaded files carry a watermark by default (official wording: Watermarked downloads by default), and watermark-free downloads require an upgrade to Pro. The reset date, what happens when the quota is used up, and whether the free plan may be used commercially are all unpublished. (Read from usevelo.ai/pricing on 2026-08-09)

See full Velo details →

Google Flow has a daily quota of 50 points (with regional restrictions; as of 2026-08-27, the official website's subscription page does not directly list this number, and it is temporarily maintained based on secondary sources), and Google AI Studio has another free quota limit.

See full Veo details →

Completely open-source and free (MIT License), self-hosted on GitHub at no cost. It comes with 18 free data sources (Eastmoney, Sina, Stooq, Yahoo, etc. that do not require an API key, and Finnhub, Alpha Vantage, Tiingo, FMP which require a free API key). Advanced data and risk calculations require purchasing QVeris credits to unlock 63+ data sources.

See full Vibe-Trading details →

The official website does not disclose the specific free quota numbers. The Free plan includes Reload credits at pay-as-you-go rates, Limited cloud agents access, Limited Warp Drive and collaboration feature access, and Limited cloud conversation storage, without pre-allocated monthly credits (Checked on 2026-08-16 from the official website)

See full Warp details →

Light daily and weekly quotas, only some models available, with unlimited Tab autocomplete and inline editing (changed from a credit system to a quota system on 2026/3/19). Re-check on 2026-08-11 found that windsurf.com/pricing now permanently redirects to devin.ai/pricing, that the page is entirely Devin-branded and that the standalone Windsurf pricing page no longer exists (the brand has been folded into Devin). The plan figures here are still consistent with the new page and have not changed.

See full Windsurf details →

Free version transcription character limit: 2,000 characters/week on desktop, 1,000 characters/week on iPhone, unlimited on Android, resets weekly. The official website states Word limit: 2,000/week on desktop, 1,000/week on iPhone, Unlimited on Android. Meeting recordings have a separate weekly limit, but the official website does not disclose the specific number, only noting that recordings under 5 minutes do not count towards the quota (checked on 2026-08-16)

See full Wispr Flow details →

Fully open source and free (MIT licence), no subscription and no platform usage cap; a CLI that runs locally. If you use it with an external LLM API, those API charges are billed separately by the model provider you choose, at that provider's list price

See full YYLO details →

Personal：2,000 次「已採納的編輯預測」(accepted edit predictions)；官網定價頁未載明此額度的重置週期。另註明自帶 API key 或外接 Claude Agent、Codex CLI 等外部代理可無限使用

See full Zed details →