# FreeBuff vs Clixad: two ad-funded coding agents, and where the ad sits > Source: > Published: 2026-08-12 00:00:00+00:00 # FreeBuff vs Clixad: two ad-funded coding agents, and where the ad sits Two free terminal coding agents, both funded by advertising, both including model access so you never supply an API key. The difference that decides everything else is **where the ad is rendered**. FreeBuff puts small text-only ads between agent turns, inside your session, and its launch post says prompts and messages may be analyzed to personalize them. Clixad puts no ad in the session at all: you complete an advertiser action in a browser tab, and that payout becomes credits. I build the second one, so read this for the sourced numbers rather than the verdict. The honest summary is that FreeBuff is the faster thing to start and Clixad is the one whose funding has no use for your prompts — and that their allowances are measured in units that cannot be converted into each other, so nobody can tell you which is "more". ## Two products that share a funding word [An earlier post here](/blog/ad-supported-ai-coding-agents/) argued that "ad-supported" is doing too much work as a label, because three different arrangements are sold under it: in-session ads, a vendor-funded daily grant, and a rewarded offerwall. That post also concluded that the in-session model had been tried by the best-resourced company in the category and abandoned inside six months, when Sourcegraph dropped ads from Amp in March 2026. That conclusion needs a correction, and this post is partly it. FreeBuff launched on **12 February 2026** with in-session text ads and is still running them in August. One vendor leaving a model does not close it, and a second implementation that has lasted six months is evidence the first read was too quick. So this is the concrete version of that earlier argument: two live products, both ad-funded, compared on what they actually grant. ## What FreeBuff grants FreeBuff is the free, ad-supported build of Codebuff, published under Apache-2.0. Everything in this section is from FreeBuff's own launch post and README, linked in [Sources](#sources-title); where two review sites disagreed with each other about the session limit, I have used the primary source and ignored both. - **No account, no card.** The launch post's words are "No account. No credit card. No config file." You install the CLI and start. - **Limited mode:**"Deepseek v4 Flash only, 6 one-hour sessions per day." This is what users outside the supported regions get, and what VPN users get. - **Full mode:** a wider model picker, described at launch as available in 25+ countries. The exact roster has moved between the launch post and the README, so check the current one rather than trusting a number in any article, including this one. - **The ads:**"small, text-only ads from sponsors" between agent turns — not inside generated code, and not as popups. That is a genuinely low-friction product. There is no signup, no wallet, no earning step and no moment where the tool asks you to go and do something else first. It is worth saying that clearly before the part where I disagree with it. ## What Clixad grants Clixad meters usage in credits against real token consumption. The rate is fixed in code: **100,000 credits is one US dollar** of advertiser payout. A new account starts with 5,000 credits, and a free account can earn at most 500,000 credits in a day. Credits convert into work directly, which is the one advantage of metering something real. The table below prices [the same twelve-call agent task](/blog/free-ai-coding-tier-limits/) used across this blog — one instruction that the agent turns into twelve model calls, resending the whole conversation each round — on every model in the catalogue, and then divides two Clixad grants by it. | Model | Credits per task | Tasks per completion | Tasks per day at the cap | |---|---|---|---| | GPT-5 Nano | 3,051 | 13 | 163 | | Gemini 2.5 Flash-Lite | 5,664 | 7 | 88 | | DeepSeek V4 Flash | 7,629 | 5 | 65 | | Gemini 3.1 Flash-Lite | 14,697 | 2 | 34 | | Kimi K2 | 32,297 | 1 | 15 | | Kimi K2.7 Code | 43,002 | 0 | 11 | | Claude Haiku 4.5 | 57,690 | 0 | 8 | | Gemini 3.6 Flash | 86,538 | 0 | 5 | | Claude Sonnet 5 | 115,380 | 0 | 4 | | GPT-5.2 | 106,632 | 0 | 4 | | Kimi K3 | 173,070 | 0 | 2 | | Claude Opus 5 | 288,450 | 0 | 1 | | GPT-5.5 | 293,850 | 0 | 1 | | GPT-5.2 Pro | 1,279,531 | 0 | 0 | The zeroes are the useful part. One completed offer at the low end does not buy one agent task on most of the catalogue, and the daily cap does not buy a single task on GPT-5.2 Pro. Those rows stay in the table for the same reason the awkward numbers stay everywhere else on this site: a comparison that drops its own bad rows is not a comparison. The other half of the honesty is the earning side. Credits arrive **on completion**, and the most common outcome of starting a survey is a **screenout**: a few questions in, it decides you are not the respondent it is paying for, and it ends with nothing credited. That is how offerwalls work industry-wide and I cannot engineer it away, because it happens on the wall's side before any postback reaches us. Observed payouts on our CPX wall in Germany, on completion, run from roughly 40,000–42,000 credits for a 3 minute action to roughly 300,000 for a 16 minute one, with a 6 minute action anywhere in 45,000–107,000. Duration barely predicts payout, which is why every figure here is a range. ## The units do not convert Here is where a comparison article normally produces a winner, and where this one cannot. FreeBuff's free allowance is measured in **time**: six one-hour sessions a day in limited mode. Clixad's is measured in **tokens**, priced into credits. There is no exchange rate between them, and inventing one would be the exact error [the tier-limits post](/blog/free-ai-coding-tier-limits/) exists to name. An hour spent reading a diff, thinking, and typing one careful prompt costs Clixad almost nothing and consumes a whole FreeBuff session. An hour of dense agent work on a frontier model is the other way round. The two allowances are not larger or smaller than one another; they are priced against different things you might run out of. What can be said without cheating: if your constraint is *wall-clock time at the keyboard*, a session limit is the thing that will bite you. If your constraint is *how much the model has to read and write*, a token meter is. Long sessions with little inference favour the meter. Short bursts of heavy inference favour the clock. ## Where the ad sits, and what it has to see This is the difference that does not come out in the wash, and it follows from geometry rather than from anybody's good intentions. An ad rendered *inside a coding session* is selected by a process that is already holding your session. The vendor then has to decide how much of it informs the selection. FreeBuff answers that question in its launch post, and the sentence is worth quoting exactly rather than characterising: "Prompts and messages may be analyzed to personalize ads, using Freebuff systems and service providers acting on our behalf." The same page draws a boundary in the other direction — "Separate uploads and connected repositories are not provided to advertising providers" — and notes that models or features labelled "May use data for AI training" may keep submissions to develop and improve AI models. So by FreeBuff's own description: prompts and messages are in scope for ad personalisation, repositories and uploads are not handed to ad providers. That is a more careful line than most products in this space draw, and I would rather quote it than paraphrase it into something worse. An ad rendered *in a separate browser tab, paying for a completed action*, has a different shape available to it. The advertiser is buying a finished survey or signup and pays a fixed amount for it; the payout arrives at our server as a signed postback naming a transaction and an amount. There is no step in that path where a prompt would help, and no step where one is present. That is not a promise about our character — [a policy is a promise and can be revised](/blog/free-ai-coding-tier-privacy/) — it is what the money path is made of. The trade is real in both directions. Clixad's arrangement asks you to leave the terminal, do something unrelated, and accept that it often pays nothing. FreeBuff's asks for none of that and instead lets the funding touch what you type. Which of those is worse depends entirely on what you are working on, and I do not think there is a general answer. ## What each one is better at **Pick FreeBuff if** you want to be coding in a minute with no account, no wallet and no interruption; if you would rather be shown a text ad than go and complete a survey; if a session-based clock fits your day better than a token meter; or if you want an Apache-2.0 codebase you can read. **Pick Clixad if** you want frontier models on the same funding as the cheap ones, with no locked tier — a stronger model simply costs more credits per turn; if you would rather no part of your prompt be involved in selecting an advertisement; or if you want an allowance denominated in the thing you actually consume, so an unused hour costs nothing. **Pick neither if** you already have an API key you are happy paying for. Both of these exist for people who do not, and a key you already hold beats both arrangements on predictability. That is not a difficult admission: the point of writing the numbers down is that they sometimes point away from you. These two are also not the only options with inference included. [The shortlist post](/blog/free-ai-coding-agent-no-api-key/) puts both alongside Copilot Free, Cursor's Hobby plan, Amp Free and the Gemini CLI free tier, and applies the same test to all six: what does the grant actually finish? ## FAQ ### What is the difference between FreeBuff and Clixad? Both are free terminal coding agents funded by advertising rather than by a subscription, and both include model access so you do not supply an API key. The difference is where the ad is rendered. FreeBuff shows small text-only ads between agent turns, inside the session, and its launch post states that prompts and messages may be analyzed to personalize ads. Clixad shows no ad in the session at all: you complete an advertiser action on an offerwall page in a browser tab, and the payout for that completed action becomes credits. Because the advertiser is buying a completed form rather than attention during work, no prompt content is involved in selecting or measuring the ad. ### Is FreeBuff or Clixad cheaper? Neither costs money, and the two allowances cannot be converted into each other. FreeBuff grants time: six one-hour sessions per day in limited mode, on DeepSeek V4 Flash. Clixad grants credits, which are metered against real token usage, so an allowance converts directly into model calls but says nothing about how many hours you sit at the terminal. An hour of thinking with no calls costs nothing on Clixad and consumes a session on FreeBuff. A dense hour of frontier-model work is the other way round. Anyone claiming one is cheaper is comparing units that do not divide. ### Does FreeBuff read my code to target ads? FreeBuff's launch post states: "Prompts and messages may be analyzed to personalize ads, using Freebuff systems and service providers acting on our behalf." It also states that "Separate uploads and connected repositories are not provided to advertising providers," and that models or features labelled "May use data for AI training" may keep submissions to develop and improve AI models. So prompts and messages are in scope for ad personalisation by their own description, while uploads and connected repositories are not passed to advertising providers. Read the current version before deciding; this is quoted from the launch post as published. ### What is a rewarded offerwall in a coding tool? It is the model mobile games have used for fifteen years. Instead of watching an ad while you work, you choose to complete an advertiser action outside the tool, usually a survey or a signup form in a browser tab, and the advertiser's payout for that completed action is converted into usage credits. The advertiser is paying for the completed action itself, so it has no reason to want your session and no route to it. The cost is friction and uncertainty: most survey attempts end in a screenout that pays nothing, which is normal across the offerwall industry rather than a fault in any one integration. ### Which should I install? If you want to start coding in the next sixty seconds with no account and no interruption, install FreeBuff: it asks for no account, no credit card and no config file, and there is no earning step between you and the model. If you want frontier models available, or you would rather no part of your prompt be involved in choosing an advertisement, Clixad is the arrangement that structurally cannot use it, at the cost of a browser tab and a survey that often pays nothing. Both are free and both install in one command, so the cheap answer is to try both against your own work. ## Sources - Freebuff, [Introducing Freebuff: the free coding agent](https://freebuff.com/blog/freebuff-launch), published 12 February 2026 and updated 22 July 2026, for the launch date, the limited-mode allowance ("Deepseek v4 Flash only, 6 one-hour sessions per day"), the 25+ country full-mode figure, the description of the ads as small text-only ads between agent turns, "No account. No credit card. No config file.", and the three quoted sentences on ad personalisation, AI training and what is not provided to advertising providers. Read 12 August 2026. - CodebuffAI, [freebuff on GitHub](https://github.com/CodebuffAI/freebuff), for the Apache-2.0 licence, the relationship to Codebuff, and the regional split between full and limited access. Read 12 August 2026. - Sourcegraph, [Amp Free Is Ad-Free](https://ampcode.com/news/amp-free-is-ad-free), for the March 2026 removal of ads from Amp referenced in the opening section. - Per-task credit figures: Clixad's own pricing engine, `packages/pricing` , over the model catalogue read from OpenRouter on 26 July 2026, recomputed by the test beside this post. The twelve-call task shape is read from the`#session` table of[the tier-limits post](/blog/free-ai-coding-tier-limits/)so the two pages cannot drift. - Credit rate, signup bonus and daily cap: `packages/pricing/src/config.ts` (`creditUsd` ,`signupBonusCredits` ,`maxAdRewardUsdPerDay` ). - Offer payouts: amounts observed on Clixad's own CPX wall in Germany, credited on completion, July 2026.