Verified free LLM API quotas, free tiers & trial credits — checked against official sources, updated weekly.
👉 Live site with full details: freeai.alexai.top 👈
Looking for a free Gemini / Qwen / GLM / Llama / DeepSeek API key, a free tier with no credit card, or a limited-time free model? This repo is the open data behind FreeAI Radar: every row links to the official pricing page, docs or the provider's official X account, and has a last-verified date.
- ✅ Official sources only — no rumours, no aggregator copy-paste
- 📅 Deadlines, extensions and "ending soon" flags for limited-time offers
- 🇨🇳 Bilingual data (EN + 简体中文), with notes for readers in mainland China
- 📦 Machine-readable:
data/offers.csv,data/offers_zh.csv,data/offers.json
⭐ Star this repo to get notified when the weekly update lands.
51 active · 14 ended · last verified 2026-10-28
| Provider | Offer / models | Free quota | Deadline | Status | Verified | Links |
|---|---|---|---|---|---|---|
| AntDigital DTMaaS (蚂蚁数科大模型服务平台) | LingDT-Decision-1.0 (lingdt-decision-1.0; 4B and 27B versions per launch coverage) | The DTMaaS model marketplace says LingDT-Decision-1.0 is "限时免费至 10 月 30 日" (free for a limited time until Oct 30) and tags it 限时免费; the listed price is ¥0 per… | Oct 30, 2026 (time and time zone not stated) | 🔴 Ending soon | 2026-10-23 | Official · Details |
| EvoMap (EvoX Agent) × MiniMax | MiniMax M3.1 Flash (inside the EvoX Agent desktop app) | EvoMap: "MiniMax M3.1 Flash. | Oct 30, 2026 (time zone and exact time not stated) | 🔴 Ending soon | 2026-10-28 | Official · Details |
| Qoder (Alibaba) | Sonus (the only model the promotional SOTA Credits can be used with) | During Qoder 1024 Community Boost (Oct 22–30, 2026, UTC+8), eligible individual users can claim a Welcome Pack of 1,024 SOTA Credits plus the community-unlocke… | Claims close Oct 30, 2026 23:59 (UTC+8); Welcome Pack Credits usable until Oct 31, 2026 23:59 (UTC+8) | 🔴 Ending soon | 2026-10-28 | Official · Details |
| DeepSeek | DeepSeek Search API, Fetch API (server-side tools) | When you use the DeepSeek API through DeepSeek Harness, Claude Code or Codex, the server-side Search and Fetch tool calls are currently not billed. | Nov 1, 2026, 00:00 Beijing time (Oct 31, 2026, 16:00 UTC) | 🔴 Ending soon | 2026-10-19 | Official · Details |
| OpenRouter / Poolside | Laguna XS 2.1, Laguna S 2.1 | Poolside's Laguna XS 2.1 and Laguna S 2.1 have :free variants on OpenRouter at $0 per token. | Oct 31, 2026 (OpenRouter expiration date; timezone not stated) | 🔴 Ending soon | 2026-10-19 | Official · Details |
| Qoder (Alibaba) | DeepSeek-Flash (DeepSeek-V4.1-Flash-1017) | Qoder's official X account says DeepSeek-V4.1-Flash-1017 was free at 0.0x credits for free and paid users, and on Oct 24 extended it: "DeepSeek-V4.1-Flash-1017… | Until Nov 1, 2026, 00:00 (UTC+8) (extended from Oct 24) | 🔴 Ending soon | 2026-10-27 | Official · Details |
| Tencent WorkBuddy (International) | DeepSeek V4.1 Flash (updated to DeepSeek-V4.1-Flash-1017 on Oct 17, 2026) | WorkBuddy International says DeepSeek V4.1 Flash stays free in WorkBuddy until October 31, 2026, and on Oct 17 said the app now runs the latest DeepSeek-V4.1-F… | Oct 31, 2026 (time and timezone not stated) | 🔴 Ending soon | 2026-10-19 | Official · Details |
| Unisound MaaS (云知声) | U2 Flash (u2-flash, 1M context) | Unisound's pricing page says U2 Flash is "限时免费" from 2026.09.30 to 2026.10.31: input, cached input and output are all ¥0 per 1M tokens during the promotion. | Oct 31, 2026 (activity period 2026.09.30–2026.10.31; time not stated, platform prices in CNY) | 🔴 Ending soon | 2026-10-23 | Official · Details |
| Qoder (Alibaba) | Qwen3.8-Flash | During the promotion, calls to Qwen3.8-Flash in Qoder International products cost 0 Credits (the Credit rate drops from 0.1x to 0.0x). | Extended past Sep 30, 2026; Qoder says it will post the end date on the page in advance | 🟠 Extended | 2026-10-19 | Official · Details |
| OpenAI Codex (paid ChatGPT plans) | Codex and ChatGPT Work usage limits | On Oct 4, 2026 OpenAI's Codex and ChatGPT lead Tibo Sottiaux (@thsottiaux) said that for the next 28 days Codex would ship either one clear improvement or a fu… | 28 days from Oct 4, 2026 (Day 13 was Oct 17, so Day 28 is about Nov 1); no fixed reset schedule | 🟢 Active | 2026-10-18 | Official · Details |
| Zhipu AI (AutoClaw) | Models available inside the AutoClaw desktop app (credits) | China site (autoclaw.zhipuai.cn): log in to AutoClaw between Oct 26 and Nov 1 to claim 100 million tokens (10,000 credits), once per account. | Nov 1, 2026 23:59 (UTC+8); claimed credits also expire then | 🟢 Active | 2026-10-25 | Official · Details |
| SCNet National Supercomputing Internet (超算互联网 MaaS) | GLM-5.3-Flash (10M-token pack; the platform may substitute another model depending on supply) | In the SCNet MaaS promotion (Aug 13 – Nov 10, 2026), users who newly register during the promotion can claim a free 10,000,000-token pack (GLM-5.3-Flash, 1024K… | Promotion ends Nov 10, 2026 23:59 (UTC+8), or earlier when the supply runs out; each pack is valid 30 days after crediting | 🟢 Active | 2026-10-25 | Official · Details |
| HeyGen (API) | HeyGen Voice Turbo Free (heygen-voice-1-turbo-free), text-to-speech | HeyGen says its Voice Turbo model is "now free in your API": requests with "model": "heygen-voice-1-turbo-free" are free and do not require a credit balance. | Free through November 30, 2026 (time zone not stated); HeyGen says it will give advance notice before the offer changes | 🟢 Active | 2026-10-23 | Official · Details |
| Qiniu Cloud (七牛云 AI 大模型推理) | Conatus Free (qiniu/conatus-free; anonymous blended model made of several top models, per Qiniu) | For 1024 Programmer's Day through Nov 30, Qiniu Cloud offers Conatus, a limited-time free blended model (qiniu/conatus-free). | Nov 30, 2026 23:59 (UTC+8). After that the model stops serving; it does not turn into a paid model. | 🟢 Active | 2026-10-25 | Official · Details |
| OpenRouter / dots-studio | dots-3-note-preview | dots-studio's dots-3-note-preview has a :free variant on OpenRouter at $0 per token. | Dec 31, 2026 (OpenRouter expiration date; timezone not stated) | 🟢 Active | 2026-10-19 | Official · Details |
| Tencent Cloud TokenHub | All language models and multimodal-understanding models on TokenHub (plus credits for Hunyuan 3D and video generation) | Tencent Cloud TokenHub gives each main account a one-time free trial pack: every language model and every multimodal-understanding model comes with 1,000,000 f… | Campaign runs until 2026-12-31; each claimed quota is valid 1 year | 🟢 Active | 2026-10-17 | Official · Details |
| Alibaba Cloud Model Studio (Qwen) | Qwen and other Model Studio models in the Singapore region (international deployment scope) | When you first activate Model Studio in the Singapore region, Alibaba Cloud gives you a separate free token quota for each eligible model (typically 1,000,000… | Quota expires 90 days after activation | 🟢 Active | 2026-10-20 | Official · Details |
| Anthropic (Claude) | Claude API, Managed Agents, Agent SDK | Paid Claude Max and Team plans now include monthly Claude API credits: $100 (Max 5x), $200 (Max 20x), or $20 per Standard seat and $100 per Premium seat on Tea… | Ongoing (monthly) | 🟢 Active | 2026-10-17 | Official · Details |
| Cerebras | All models in Cerebras Shared Inference (e.g. gpt-oss-120b, qwen-3.8-27b) | New accounts get $5 in free credits after adding a verified payment method. | Credits expire 30 days after they're granted | 🟢 Active | 2026-10-20 | Official · Details |
| Cline | Free list on Oct 23, 2026: Reflection Beam (501B), Nemotron 3.5 Super VL (120B), Poolside Laguna S 2.1 (118B), Nemotron 3.5 Lightning (32B), Solar Mini 4 (35B) (reflection/beam-free, cline-free/nemotron-3.5-super-vl, poolside/laguna-s-2.1-free, nvidia/nemotron-3.5-lightning-free, cline-free/solar-mini4) | Cline periodically offers free model promotions that let you try select models at no cost, up to a limited usage quota; free models are available to any user w… | Not published (rotating, limited-time promotions) | 🟢 Active | 2026-10-23 | Official · Details |
| DigitalOcean (Managed Agents / Agent Droplets) | Models hosted on DigitalOcean, including Kimi K3 and GLM 5.3 (inference inside Managed Agents) | DigitalOcean Managed Agents (public preview) has a Free Trial for new customers: $5 in credit, no card required. | Not published | 🟢 Active | 2026-10-25 | Official · Details |
| Dolphin AI (Dolphin API) | Dolphin API models (including DeepSeek-V4.1-Flash-1017 per @dphnAI) | Dolphin's API page says: use code API10 for $10 in Dolphin API credit. | Not published (limited to the next 5,000 signups) | 🟢 Active | 2026-10-19 | Official · Details |
| Huawei Cloud CodeArts (码道 CodeArts agent) | deepseek-v4.1-flash, glm-5.3-flash, deepseek-v4-flash-0731, deepseek-v4-pro-0813 | Huawei Cloud's CodeArts (码道) coding agent gives every user 10 million free tokens per day during a limited-time promotion, for the promotion models deepseek-v4… | Until further notice ("另行通知") | 🟢 Active | 2026-10-19 | Official · Details |
| iFlytek Xingchen MaaS (讯飞星辰 MaaS) | Spark-X2.6 (spark-x2.6, 293B-A30B MoE, 1M context); Spark-X2.6-4B (spark-x2.6-4b) | The Xingchen MaaS homepage says "Spark-X2.6赋能开发者 零门槛领取10亿tokens免费额度" (1 billion free tokens for Spark-X2.6). | Claim deadline not published; per user reports the quota is valid for 31 days after activation. Spark-X2.6-4B free period end not published. | 🟢 Active | 2026-10-23 | Official · Details |
| Inception | Mercury 3 (mercury-3) and other Mercury models (credit shared across models) | Inception's docs say every new account includes a one-time credit of 100 million free tokens, shared across all models, with no payment details required. | Not published | 🟢 Active | 2026-10-21 | Official · Details |
| Kilo Code (Kilo Gateway) | Free on Oct 19, 2026: Nemotron 3.5 Super VL, Nemotron 3 Ultra, Nemotron 3 Super, Dots3-Note Preview, Laguna S 2.1; plus Auto Free (kilo-auto/free) | Kilo Code supports free models for coding and background tasks through the Kilo Gateway and partner providers. | Not published (free models change over time) | 🟢 Active | 2026-10-19 | Official · Details |
| KRAFTON AI | Raon-1 mini preview (30B MoE, 4B active) | KRAFTON released a preview of Raon-1 mini on Oct 21, 2026 and says: "To celebrate the model's release, API access is currently free for a limited time." Open w… | Not published ("for a limited time") | 🟢 Active | 2026-10-21 | Official · Details |
| Liquid AI (Liquid Console) | Liquid models served via API, e.g. d1, d1-ultra, LFM2.5-24B-A2B (per the pricing page) | Liquid AI pricing page: "Every new organization gets $2 in free credit." ("$2 in free credit for every new organization — Start free on Liquid Console"). | Not published | 🟢 Active | 2026-10-28 | Official · Details |
| Meituan LongCat | LongCat-2.5-Preview | Every existing LongCat API platform user gets 5 million free tokens to try LongCat-2.5-Preview. | Not published | 🟢 Active | 2026-10-20 | Official · Details |
| Morph (morphllm.com) | All Morph models, including Morph v3 (Morph API also serves Kimi K3, GLM-5.3, GLM-5.3-Flash, DeepSeek V4.1 Flash and others) | Morph: "Full access to Morph for a year, no credit card required." The students page lists "Full API access for 12 months" and "All models including Morph v3". | Not published (12 months of access after verification) | 🟢 Active | 2026-10-28 | Official · Details |
| Nous Research (Nous Portal) / Cline | Upstage Solar Mini 4 | Upstage's Solar Mini 4 (35B total / 3B active parameters, 512K context) is free on Nous Portal: Nous announced it as free for two weeks on Oct 5, 2026, and on… | Nous Portal: no end date published ("two weeks" from the Oct 5 post, but still listed at $0.00/1M on Oct 21, 2026). Cline: no end date announced; still on Cline's free list on Oct 21, 2026. | 🟢 Active | 2026-10-21 | Official · Details |
| Nscale | Nscale Serverless Inference models, including GLM 5.3 (Z.ai), Qwen3.8-27B, Qwen3.8-Flash-Next-FP8 | Nscale: "every new user can claim $5 of free credit to explore our models." The credit is used on Serverless Inference, which now lists GLM 5.3 ($1.40 / $4.40… | Not published | 🟢 Active | 2026-10-27 | Official · Details |
| NVIDIA via OpenRouter / OpenCode / Cline / Kilo Code | Nemotron 3.5 Super VL (120B total / 12B active) | NVIDIA's new open multimodal model Nemotron 3.5 Super VL (120B total / 12B active) has a $0 endpoint on OpenRouter (nvidia/nemotron-3.5-super-vl:free). | Not published | 🟢 Active | 2026-10-19 | Official · Details |
| Ollama (Ollama Cloud) / Reflection AI | Beam (beam:cloud) | Reflection's first open-weight model Beam (about 501B total / 23B active parameters, 1M-token context, Apache 2.0) is free on Ollama Cloud for a limited time. | Not published ("for a limited time") | 🟢 Active | 2026-10-19 | Official · Details |
| OpenCode (OpenCode Console, formerly Zen) | DeepSeek V4.1 Flash Free, GLM-5.3-Flash Free, Beam Free, Flint 40B Free, LongCat 2.5 Preview Free, Exo Free, MiMo-V2.6-Flash Free, MiMo-V2.5 Free, Nemotron 3 Ultra Free, Nemotron 3.5 Lightning Free, Nemotron 3.5 Super VL Free, Big Pickle, Jev 1.13 Free | OpenCode's Console pricing table (the old Zen docs URL now redirects there) lists a set of models at $0 for input, output and cached reads, including DeepSeek… | Not published ("for a limited time"; no end date in the docs or the posts) | 🟢 Active | 2026-10-23 | Official · Details |
| Qoder (Alibaba) | Qoder Credits (usable across Qoder International products) | Since Sep 18, 2026 at 10:00 (UTC+8), individual Qoder International users can claim 100 Credits once a day in the Qoder desktop app. | To be announced (no end date on the official page) | 🟢 Active | 2026-10-21 | Official · Details |
| Qoder (Alibaba) | Efficient model tier (Qoder's standard-reasoning tier) | Since Sep 3, 2026 14:00 (SGT), paid Qoder users can select the Efficient tier without spending Credits: its billing multiplier drops from 0.3x to 0.0x. | To be determined; Qoder says it will give advance notice and update the page before it ends | 🟢 Active | 2026-10-21 | Official · Details |
| Recursive (via OpenRouter, OpenCode) | Recursive Flint (40B MoE, about 2.5B active; recursive/flint-v1.0:free on OpenRouter, flint-40b-free on OpenCode) | Recursive released Flint, its first model, on Oct 21, 2026: an open-weight 40B MoE coding model (about 2.5B active parameters, 262K context on OpenRouter). | Not published (OpenRouter: "free to use at launch", no expiration date in its models API on Oct 22; OpenCode: "for a limited time") | 🟢 Active | 2026-10-22 | Official · Details |
| Shanghai AI Laboratory (Intern InkStone / 书生·端砚) | Limited-time free: Intern-S2 (intern-s2), Atria-Dawn, Agents-A1, Sage-A1.5. Paid in 墨点 (ink points): DeepSeek-V4.1-Flash-1017, DeepSeek-V4-Pro-0813, GLM-5.3, Kimi-K2.6, MiniMax-M3, Qwen3.8-27B, Intern-Decision-4B | The Token Plan model list marks Intern-S2, Atria-Dawn, Agents-A1 and Sage-A1.5 as "限时免费" (limited-time free): 0 墨点 per 1M tokens for cached input, uncached inp… | Not published (limited-time free; test period) | 🟢 Active | 2026-10-28 | Official · Details |
| Vercel AI Gateway / Reflection AI | Reflection Beam 1.0 (reflection/beam, reflection/beam-free) | Vercel AI Gateway says Reflection AI's Beam is free to use for a limited time during the launch period when serviced by Reflection AI. | Not published ("for a limited time during the launch period") | 🟢 Active | 2026-10-19 | Official · Details |
| Xiaomi MiMo Open Platform | mimo-v2.5-tts, mimo-v2.5-tts-voiceclone, mimo-v2.5-tts-voicedesign (text-to-speech) | Xiaomi MiMo's pay-as-you-go pricing page lists the TTS series as "限时免费" (free for a limited time): mimo-v2.5-tts, mimo-v2.5-tts-voiceclone and mimo-v2.5-tts-vo… | Not published ("限时免费"; no end date on the pricing pages, which were updated Oct 21, 2026) | 🟢 Active | 2026-10-23 | Official · Details |
| Z.ai (Zhipu) | GLM family | Approved startups get free Z.ai API credits of up to 1B tokens plus volume pricing on GLM, priority support and rate limits, and early API access. | No deadline published | 🟢 Active | 2026-10-12 | Official · Details |
| Zhipu AI / Z.ai (ZCode) | GLM-5.3-Flash (inside the ZCode client, Global Building) | @zcode_ai: "300M free GLM-5.3-Flash tokens on ZCode. | Not published (announced as daily) | 🟢 Active | 2026-10-27 | Official · Details |
| Cloudflare Workers AI | Llama, Gemma, Qwen, GLM-4.7-Flash, gpt-oss, FLUX image models, Whisper and more (per-model Neuron prices) | Every account gets 10,000 Neurons per day at no charge, on both the Workers Free and Workers Paid plans. | Ongoing (no end date published) | 🟢 Ongoing | 2026-10-28 | Official · Details |
| Google (Gemini API / AI Studio) | Gemini 3.8 Flash, 3.6 Flash, 3.5 Flash-Lite, 3.1 Flash-Lite, 3 Flash Preview, 2.5 Pro, 2.5 Flash, 2.5 Flash-Lite, Gemini Embedding 2, some Live and TTS models | The Gemini API Free tier charges nothing for input and output tokens on the models marked "Free of charge" on Google's pricing page, including Gemini 3.8 Flash… | Ongoing (no end date published) | 🟢 Ongoing | 2026-10-20 | Official · Details |
| Groq | Models on GroqCloud (e.g. openai/gpt-oss-120b, gpt-oss-20b, qwen/qwen3.8-27b, Whisper) | GroqCloud has a Free tier: you can create an API key and call models without adding a payment method. | Ongoing (no end date published) | 🟢 Ongoing | 2026-10-20 | Official · Details |
| Mistral AI | Mistral models on the API (e.g. mistral-small-latest) | New Mistral accounts start in Free mode, with API access on by default and no credit card required. | Ongoing (no end date published) | 🟢 Ongoing | 2026-10-20 | Official · Details |
| Nous Research (Nous Portal) | Models priced $0.00/1M on Oct 19, 2026: Reflection Beam 1.0 501B A23B, StepFun Step 5 Preview, StepFun Step 3.7 Flash, Upstage Solar Mini 4, Meituan LongCat 2.0, Meituan LongCat 2.5 Preview, Poolside Laguna S 2.1, Poolside Laguna XS 2.1 | Nous Portal has a $0 Free plan: "Free models only", "Standard rate limits", "$0 monthly credits". | Ongoing plan (individual free models come and go) | 🟢 Ongoing | 2026-10-19 | Official · Details |
| NVIDIA (NIM / build.nvidia.com) | Models in the NVIDIA API catalog at build.nvidia.com | Members of the free NVIDIA Developer Program get free access to NIM API endpoints for prototyping, and can download NIM microservices for research, development… | Ongoing while you're a program member | 🟢 Ongoing | 2026-10-20 | Official · Details |
| OpenRouter | Models whose ID ends in :free (the list changes often) | OpenRouter offers free variants of some models; their IDs end in ":free". | Ongoing (individual free models come and go) | 🟢 Ongoing | 2026-10-19 | Official · Details |
| Zhipu AI / Z.ai (GLM) | GLM-4.7-Flash, GLM-4.5-Flash, GLM-4.6V-Flash (Z.ai); also GLM-4-Flash-250414, GLM-Z1-Flash, GLM-4V-Flash, GLM-4.1V-Thinking-Flash, CogView-3-Flash, CogVideoX-Flash (BigModel China) | Zhipu's international platform Z.ai lists GLM-4.7-Flash, GLM-4.5-Flash and GLM-4.6V-Flash as Free for input, cached input and output. | Ongoing (no end date published) | 🟢 Ongoing | 2026-10-20 | Official · Details |
| Provider | Offer / models | Free quota | Deadline | Status | Verified | Links |
|---|---|---|---|---|---|---|
| Ant Ling / OpenRouter | Ling-3.0-flash-VL | Two weeks of free access on OpenRouter. | Ended Sep 23, 2026, 9:00 AM PT; now paid at discounted prices | ⚫ Ended | 2026-10-09 | Official · Details |
| Cline | DeepSeek-V4.1-Flash | Free DeepSeek-V4.1-Flash in the Cline provider. | Paused Oct 4, 2026 because of abuse | ⚫ Ended | 2026-10-18 | Official · Details |
| GMI Cloud (with Alibaba Qwen) | Qwen 3.8 Max, Qwen 3.8 Flash, Wan3.0-video | Free use of Qwen 3.8 Max, Qwen 3.8 Flash and Wan3.0-video through GMI Cloud's OpenAI-compatible API during the promotion. | Paused early on Oct 11, 2026 (GMI announcement at 01:58 UTC / 09:58 Beijing) due to a traffic surge | ⚫ Ended | 2026-10-12 | Official · Details |
| Ant Ling via Novita AI | Ling-3.1-Flash | Free API access to Ling-3.1-Flash on Novita AI (OpenAI-compatible, with Anthropic API support). | Oct 13, 2026, 9:00 AM PT (Oct 14, 00:00 UTC+8) | ⚫ Ended | 2026-10-17 | Official · Details |
| Apodex via Novita AI | Apodex 1.1 Mini | Free trial of Apodex 1.1 Mini (35B parameters, 262K context, native tool/function calling) on Novita AI, launched Oct 1, 2026. | About Oct 15, 2026 on Novita (14-day trial from Oct 1; exact end not published). Concentrate AI: Oct 14, 00:00 Singapore time (UTC+8). | ⚫ Ended | 2026-10-17 | Official · Details |
| Cognition (Devin) | SWE-2 | Cognition's SWE-2 coding model is free in Devin Desktop and Devin CLI for self-serve paid plans (Pro $20/month and up) during the launch period. | Pricing page: through Oct 16, 2026; model docs: through Oct 15, 2026 (timezone not stated) | ⚫ Ended | 2026-10-17 | Official · Details |
| StepFun (阶跃星辰) | Step 5 Preview | StepFun announced a week of free access to Step 5 Preview starting Oct 8, 2026 in coding tools including OpenCode, Cline, Nous Research (Nous Portal) and Kilo… | About Oct 15–16, 2026 (one week from each tool's launch on Oct 8–9; exact end not published) | ⚫ Ended | 2026-10-17 | Official · Details |
| Anthropic Claude (Pro, Max, Team) | Claude app and Claude Code usage limits | With the Claude Opus 5.5 launch on Sep 22, 2026, Pro, Max and Team subscribers got one saved limit reset. | Oct 22, 2026 (per @ClaudeDevs; your exact expiry is shown in Settings → Usage) | ⚫ Ended | 2026-10-23 | Official · Details |
| MiniMax (free period ended on all channels: 0G Private Computer, Vercel AI Gateway, OpenRouter, OpenCode, Ollama Cloud, Cline, Kilo) | MiniMax M3.1 Flash (minimax/minimax-m3.1-flash; previously the stealth model Space Bunny) | MiniMax released M3.1 Flash on Oct 19, 2026 and confirmed the stealth model Space Bunny was an earlier version. | Ended. Last channel, 0G Private Computer, ended early on Oct 23, 2026 16:00 UTC (Oct 24 00:00 UTC+8) instead of Oct 24 13:00 UTC. | ⚫ Ended | 2026-10-24 | Official · Details |
| Zhipu AI (AutoClaw) | Models available inside the AutoClaw desktop app (credits) | China site (autoclaw.zhipuai.cn), v2.1.0 upgrade perk: log in on any day from Oct 15 to Oct 21 to claim 200 million tokens (20,000 credits) once per account; a… | China: one-time pack Oct 21, daily pack Oct 25; International: Oct 25 (UTC+8) | ⚫ Ended | 2026-10-25 | Official · Details |
| Zhipu AI / Z.ai (ZCode) | GLM-5.3 and GLM-5.3-Flash (inside the ZCode client) | For 1024 Programmer's Day, ZCode is giving away 200,000 packs of 500M free tokens: GLM-5.3 (200M) + GLM-5.3-Flash (300M), per @zcode_ai. | Oct 26, 2026 01:00 UTC (09:00 UTC+8), or when the 200K packs run out | ⚫ Ended | 2026-10-26 | Official · Details |
| Baseten (Model APIs) / Vercel AI Gateway | mammoth-1 (27B) and elephant-1 (9B) decision models from Base Labs (baseten/mammoth-1, baseten/elephant-1) | Baseten released two open-source decision models, mammoth-1 (27B) and elephant-1 (9B), on Oct 21, 2026 and says both are "free to try for a week on Baseten Mod… | Ended. Baseten: the free week ended Oct 27, 2026 at 11:59 PM PT (Oct 28, 06:59 UTC; Oct 28, 14:59 Beijing time, UTC+8; per @baseten on X). Vercel AI Gateway: through Oct 27 (time zone not stated). | ⚫ Ended | 2026-10-28 | Official · Details |
| Nous Research (Nous Portal) | MISSINGNO. (stealth model) | MISSINGNO., a token-efficient coding and agentic reasoning model, is free for a limited time on Nous Portal and can be used in Hermes Agent. | Ended by Oct 19, 2026 (no longer in the Nous Portal model catalog; no end announcement found) | ⚫ Ended | 2026-10-19 | Official · Details |
| OpenRouter / Inception | Mercury Decide | Inception's Mercury Decide has a free endpoint on OpenRouter, next to a new paid zero-data-retention endpoint ($0.02/M input; output and cached input free). | Free endpoint no longer listed (checked Oct 12, 2026) | ⚫ Ended | 2026-10-12 | Official · Details |
Full conditions (rate limits, card / phone requirements, regions, commercial use, step-by-step claim guide) are on each offer's page at freeai.alexai.top.
| Badge | Meaning |
|---|---|
| 🟢 Ongoing | Permanent free tier, no end date published |
| 🟢 Active | Running promotion or new-user credit |
| 🟠 Extended | The provider officially extended the deadline |
| 🔴 Ending soon | Ends within 3 days |
| ⚫ Ended | Expired or paused; kept for reference |
Dates are as published by the provider; when the official page gives no number we write "Not published" instead of guessing.
| File | What |
|---|---|
data/offers.json |
Source of truth: mirror of the site's public feed freeai.alexai.top/offers.json (EN + ZH per offer) |
data/offers.csv |
One row per offer: structured fields + English text (generated from the JSON) |
data/offers_zh.csv |
Simplified Chinese text for the same slug (generated from the JSON) |
Key columns: status (active/expired), category (evergreen, limited, trial_credits, subscriber), end_date (YYYY-MM-DD, drives auto-expiry), card_required, region, source_url (official), last_verified. Everything here is synced weekly with python3 scripts/gen_readme.py --fetch, which downloads the feed and regenerates the CSVs and the tables above.
Found a new free quota, a changed limit or an ended offer? Open an issue — with a link to an official source (provider pricing/docs page, official blog, or the provider's official X account). Claims backed only by third-party posts, KOLs or aggregator lists are not accepted. See CONTRIBUTING.md.
Free quotas change fast and can end without notice. This list is for information only and is not affiliated with any provider — always read the provider's official terms before relying on an offer. Use one account per person as the terms require: multi-accounting, bulk sign-ups, farming or reselling free quotas violate most providers' terms and get accounts banned, and this project does not cover or support them.
MIT for code and data compilation. Product names and trademarks belong to their owners.