AI tool deals, discounts & free credits
Browse verified AI deals, subscription discounts, promo codes, free trials, and API credit grants with eligibility rules and expiration status.
245 deals · page 2 of 11
AMD has launched Token Lab in its global AI Developer Program. Join for free and you get an API key with 10 credits to spend over 30 days on open models such as DeepSeek-V4.1-Flash, Qwen3.8-27B and Gemma-4-31B-IT, through an OpenAI-compatible endpoint.
Higgsfield’s 11.11 annual sale of up to 60% off ended at 00:00 UTC on November 16 (8:00 AM Beijing time). During the sale, Plus worked out to $35 a month billed yearly, and Plus or Ultra also included 7 days of unlimited generation.
Fish Audio has opened its flagship S2.1 Pro voice model to developers for free. Set the model to s2.1-pro-free in your API calls and there's no hard character cap under fair use. The free window runs through November 30.
Nex-N2.6, an open-weight multimodal agentic model post-trained on GLM-5.3-Flash, is free for a limited time on MOSS API. Create a MOSS API key and every request costs 0 Credits. It takes text, image and video input, works with both OpenAI and Anthropic API formats, and no end date has been announced yet.
JetBrains now gives every JetBrains Account free access to Junie Lite, the free tier of its Junie coding agent. It currently runs on Qwen 3.6 27B, everyday use is virtually unlimited, and no subscription or card is needed. Sign in through Air in IDE 2026.3 or Junie CLI to start.
The 48-hour free Gemini 4 Carbon route on Respan has ended. The model page now lists paid routes only, currently 50% off through about December 1, 2026. To keep calling the model, you need a paid route.
Qianwen Go!'s Double 11 membership sale has ended. The half-off first month closed at midnight at the end of November 13, Beijing time, and new sign-ups now pay the regular price.
Claude paid plans get one free full usage reset
Anthropic gave every Claude Pro, Max, Ultra, and Team user one free full reset across all models and limits. You can use it anytime before December 10 at 2:00 pm PT.
Meoo is offering new-user first-month pricing of ¥25 per month for Lite (regularly ¥29), ¥39 for Pro (regularly ¥59), and ¥89 for Max (regularly ¥179). The ¥25 tier is Lite, not Pro. The next billing cycle returns to the regular price. New accounts also receive 5,000 Credits, which we already cover on a separate page.
Volcengine Ark Doubao flagship packs are ¥19 for about 4.4 million tokens!
On Volcengine's 11.11 AI campaign, a Doubao flagship model pack costs ¥19 for about 4.4 million tokens, down from ¥38. The same page also advertises first-order flagship packs from 50% off, and a limited 75% off first-order offer on Ark Agent Plan. The campaign runs through 23:59:59 on November 30, China Standard Time. The annual Agent Plan half-off sale and the 500,000 free tokens per model are separate offers we already published.
Baidu Qianfan Token Plan is up to 50% off!
Baidu AI Cloud's Qianfan 11.11 campaign offers Token Plan at up to 50% off, and the same page also lists the enterprise Token Plan at up to 50% off. The sale runs through November 30, 2026, and you need a registered, identity-verified Baidu AI Cloud account. The campaign page does not list a discounted price for each tier, so confirm the amount at checkout.
Personal general or Hy Token Plan subscribers who call Hy4 from November 4 through 12:00 on November 17 (China Standard Time) are billed at 25% of the usual Credit price: 30 Credits (¥1.5) per million input tokens on a cache miss, 90 Credits (¥4.5) per million output tokens, and 1.5 Credits (¥0.075) on a cache hit. Pay-as-you-go Hy4 usage is billed at 50% of the list price from the date the notice was published through 23:59:59 on December 31, against list prices of ¥6 / ¥18 / ¥0.3 per million tokens.
On the Qwen AI Double 11 campaign page, first-time personal Token Plan purchases on Alibaba Cloud Model Studio (Bailian) are discounted by up to 50%. The campaign lists monthly allowances of 20,000 Credits for Lite, 45,000 for Essential, and 80,000 for Standard. For pay-as-you-go usage, the AI savings plan is as low as 55% off on a quarterly commitment and covers 150+ first-party models. The campaign runs through November 30.
The TRAE Bao member free window ended at 23:59:59 on November 15, 2026. During the offer, members could use up to 1,000 free queries in the desktop app.
TRAE China annual membership is 30% off!
TRAE China's Double 11 membership sale runs from November 3 to November 16. Annual plans are 30% off, and continuous monthly plans are 10% off, with no coupon or bundle required. Lite, Pro, Pro+, and Ultra are all included. The pricing page currently shows 'as low as' monthly figures of ¥34, ¥69, ¥167, and ¥489, against regular monthly prices of ¥49, ¥99, ¥239, and ¥699.
MiniMax Code gives 10 free H3 Image generations
Sign in to MiniMax Code on the web, desktop, or mobile app and generate images with H3 Image. Everyone gets 10 free generations at 1K, 2K, or 4K. The offer runs through November 17; after the free images are used up, further generations use account credits, and 4K output is still billed at half credits.
MiniMax is taking 50% off 4K image generation with its new MiniMax-H3-Image model: when the delivered image is at least 16.6 megapixels, the whole request is half price, about $0.034 per standard 4K text-to-image. The offer runs until November 17, 2026 at 16:00 UTC, and 1K and 2K images stay at regular price.
Google's most capable reasoning model, Gemini 4 Carbon, is half price on the paid API through November 30, 2026: $2 per million input tokens and $10 per million output tokens, going back to $4 and $20 on December 1. You can call it at this price on the paid Gemini API or on Vercel AI Gateway; the Free tier doesn't include this model.
SCNet Gives New Users 100M Free Tokens (GLM-5.3-Flash)
New SCNet (China's national supercomputing internet platform) users can claim 100 million free tokens from the "New User" entry in the desktop client. The default model is GLM-5.3-Flash, valid for 30 days after credit. Runs through December 31, one claim per new account.
Claude Fable 5.2 is out, and from November 9 to November 16 Pro plans and standard Team seats can run Fable on their regular plan limits, up to half of weekly usage, plus one free Fable reset. Max and Ultra subscribers get one to three Fable resets based on how long they've subscribed.
New ZCode users get a free daily GLM trial quota with no upfront payment: connecting BigModel grants ~8M tokens/day for 5 days (3M GLM-5.3 + 5M GLM-5.3-Flash), and overseas users connecting Z.ai also receive a free daily GLM trial quota.
Rime's free plan needs no card on file and renews about $50 of usage every month for Prism conversations, Mist v3 / Coda v2 TTS, and Rime STT.
OpenCode Zen added a free tier for Meituan LongCat 2.5, the official release. Input, output, and cache reads cost nothing, with a limit of 100 requests per day. It supports about one million tokens of context and image input. The provider keeps zero data and does not train on your requests. Free chat models must be used through OpenCode, and no end date has been announced.

