A request is not a unit of work: what free AI coding limits actually buy

Every comparison of free AI coding tools ends up as a table of limits: so many requests a month, so many premium requests, a daily grant that resets at midnight. None of those numbers tells you how many jobs you can finish before the tool stops.

Measuring that got harder when coding assistants became agents, and the units on the pricing pages never caught up.

Five units, none of them work

Request

One thing you asked for. GitHub's documentation defines it as "any interaction where you ask Copilot to do something for you", and states that for agentic features only the prompts you send are counted, while the tool calls the agent makes on its own are not. The quota therefore measures your typing, not the model's work.

Premium request

A request weighted by a per-model multiplier. The multiplier exists because a request has no fixed size: it corrects for the model you picked, but not for the length of the job. GitHub documents one Copilot code review as consuming 13 premium requests.

Daily grant

An allowance of model usage that refreshes each day, usually stated in a currency the vendor does not publish a conversion rate for. It is the only unit here that scales with real cost rather than with interactions, and also the one most easily reduced, because it is set by whoever is funding it. Amp cut its free daily allowance for more users in May 2026.

Token

The unit providers actually bill in. It is exact and it is the only one that survives contact with an agent loop. A raw token count still hides the price difference between models, though, where the same token can cost two hundred times more depending on which one you sent it to.

Credit

Tokens already priced at the model's rate, so one number covers both how much work you did and which model did it. Clixad meters in credits, at 100,000 credits to $1 of list value. We chose this unit and are not neutral about it. A credit is only as useful as the published conversion rate behind it.

One task, twelve calls

Take an instruction an agent gets a hundred times a day: find why this test fails and fix it. The agent reads the failing test, greps for the function, reads two files, makes an edit, runs the suite, reads the new error, edits again, runs it again, and reports back. On a request-counted plan that is one request. In the provider's logs it is roughly a dozen calls.

A model has no memory between calls, so every call carries the entire conversation so far. Each file the agent reads and each command it runs is appended and sent again on every later call. The context accumulates in your bill.

Here is the task written out. It starts at 12,000 input tokens, which is a system prompt, a project file and your instruction. Every step adds 3,100 tokens of tool output and previous reply, and every step generates 600 tokens back. Twelve steps.

The twelfth call sends 46,100 tokens to answer a question you asked once. Cumulatively the task costs 348,600 input tokens and 7,200 output tokens. If you had estimated it as twelve calls the size of the first one, you would have said 144,000, and you would have been under by a factor of about two and a half.

This is why "how many requests do I get" has no useful answer. The same request is a tenth of the bill on a one-shot question and forty times that on a debugging session.

What that task costs on fourteen models

Below is the same twelve-step task priced through the engine Clixad meters with, on every model in our catalogue. Each call is metered and rounded on its own, the way the gateway actually bills it. The second column is the estimate you would have reached by pricing twelve calls at the size of the first.

The undercount runs between roughly two and two and a third times across every model, so it is not a quirk of one price list. And the spread from the top row to the bottom one is nearly two hundred times, for the same task.

The cheapest model depends on the shape of your work

Our model catalogue is ordered by what a single conversational turn costs, defined in the previous post as 20,000 input tokens and 1,500 output tokens. By that measure Gemini 2.5 Flash-Lite costs 390 credits and DeepSeek V4 Flash costs 484, and the twelve-step task above agrees: 5,664 against 7,629.

Ask for output instead and the order reverses. A call that sends 2,000 tokens and gets 2,000 back, which is the shape of asking for a file to be written, costs 150 credits on Gemini 2.5 Flash-Lite and 126 on DeepSeek V4 Flash. Nothing changed about the models. Flash-Lite charges less for input and more for output, both of those turns are input-heavy, and the writing call is not.

So there is no single price ranking to memorise, only one per workload shape. A league table of models by price shows the shape its author had in mind.

What one offerwall completion covers

Clixad funds usage with rewarded offerwall completions, so the same conversion applies to us. From our own ledger: completed offers on our CPX wall have credited from 9,000 to 114,000 credits on completion, with a median of 51,000: 26 completions recorded in our ledger between 27 July 2026 and 12 September 2026, as of 14 September 2026.

The largest paid more than twelve times the smallest, and the ledger does not record how long an offer took. Take the median, 51,000 credits, against the table above: it covers about nine of these tasks on Gemini 2.5 Flash-Lite, about one and a half on Kimi K2.5, and about a fifth of one on Claude Opus 5.5.

Being screened out is the normal case, not an error. The most common outcome of starting a survey is that it ends part way through and pays nothing or a small bonus, because you did not match the panel being bought. That is how offerwalls work everywhere.

We also cap what one free account can earn per day, at 500,000 credits.

A completion is not the only shape ad funding takes, either. The other live one is funded by the advertisement inside the session rather than by an action you complete elsewhere, which changes what the funding needs to see. A later post compares the two directly.

Converting any quota into work

You can do this arithmetic on any tool, including ours, from public pages and an afternoon of use. A later post runs it across the six free tiers that include model access and finds that four of them grant a unit it cannot be applied to. Others apply it to a $5 trial balance, to the sources of free credits, to Gemini CLI's daily request quota, and to when each of those tiers resets.

1. Find out which model is actually serving you on the free tier. Not which models the vendor sells. Free tiers get restricted to the cheap end first, which Google did to the Gemini CLI free tier in March 2026, and which is one of the defaults a vendor can move without touching a single published limit.
2. Price one task, not one message. Give the agent a job you would really give it and count the model calls it makes. A dozen for something ordinary is a reasonable planning number.
3. Multiply the context growth back in. If you estimate from the first call, double it. That is the correction the table above measures.
4. Convert the quota into that unit. If you cannot, because the quota is in requests and requests have no size, you have learned the thing worth knowing about the offer.
5. Ask what happens at zero. A tier that degrades to a smaller model is a different product from one that stops, and both are different from one that silently truncates your context to fit. A later post takes that question apart, including what an abandoned task has already cost by the time it stops.

What Clixad does, and where it also runs out

Clixad is a terminal coding agent funded by rewarded offerwall completions, a sponsored line in the terminal and one-time credit packs rather than by a subscription. It meters in credits at a published rate, it shows you the price of a turn before you spend it, and the wallet is checked before generation starts rather than after, so a request you cannot afford is refused instead of being half-served and clawed back. New accounts start with 25,000 credits.

By the numbers above, that is four full twelve-step tasks on Gemini 2.5 Flash-Lite, and it does not cover one full twelve-step task on Kimi K2.5, from the same cheap end of the catalogue. What a signup bonus is worth depends on the model you spend it on. Under the daily earning cap, the most a free account can draw in a day on Claude Opus 5.5 is two tasks of the kind priced here, and not a third.

What it does fix is the unit: you can decide whether it is worth a survey from published numbers.

Later posts price this task against Claude Code, the Codex CLI and FreeBuff, and each says who should pay for the other tool instead.

FAQ