Is the Claude API Free? Real Rate Limits & Costs in 2026
Last verified: 31 October 2026, checked against Anthropic’s pricing, rate-limit and support docs.
No. The Claude API isn’t free, at least not in any ongoing way. New Claude Console accounts get a small bit of trial credit to kick the tires, and after that it’s pay-as-you-go — you prepay credits and get billed per million tokens. There’s no permanent free tier, nothing like the always-free daily allowances some other APIs offer. The better news is that it’s cheap for light use, and there are a few legitimate ways to keep your bill near zero. Here’s how it actually works in 2026.
What changed in September and October. Claude Sonnet 5.5 launched on 28 September at the same $2/$10 as Sonnet 5, and Sonnet 5 moved to Anthropic’s legacy list. Claude Opus 5.5 arrived the week before at $4/$20, cheaper than Opus 5. Sonnet 4.5 now has a retirement date, 30 November. In October, Max 20x was renamed Claude Ultra at the same $200; Anthropic lists it at 10x Pro’s usage every week, with no five-hour limit for verified accounts. Sonnet 5.5 also got higher request limits: 4,000/10,000/13,000 RPM at Start/Build/Scale, while Sonnet 5 stays at 1,000/5,000/10,000. Since the May version, Opus has stopped being the top of the price ladder and the usage tiers have been renamed. An earlier version said Sonnet 5’s price would rise on 1 September; Anthropic cancelled that rise.
Prices move; the date at the top is when these were last checked.
The short answer
Free forever? No. Free to start? Yes — new Console accounts get a one-time-ish trial credit, enough to build a prototype, not enough to run a product. After that, it’s pay-as-you-go: prepay credits, billed per million input and output tokens, with the price depending on which model you call.
Is it cheap? For light use, yes. A short completion is a fraction of a cent. For heavy agentic work — big context, long outputs — it adds up fast.
And if what you actually want is “Claude, but flat-rate instead of metered,” you probably want a subscription, not the API: Claude Pro, Max or Ultra, or Claude Code running on it. That’s a fixed monthly fee with a usage allotment, which is what a lot of people mean when they search “is the Claude API free.”
What “free” actually means here
When people ask whether the Claude API is free, they’re usually asking one of a few different things. Honest breakdown:
The new-account trial credit. Create an account in the Claude Console and you get a small amount of credit to play with. It’s real, it’s free, and it’s the closest thing to a free tier — but it’s a one-time grant, not a recurring quota. Treat it as enough to validate an idea, not run one.
Prompt caching and the Batch API. Not “free,” but the biggest lever you’ve got on cost. A cache read costs 10% of the standard input price — 90% off — on the parts of your prompt that repeat across calls: system prompts, long context, few-shot examples. Writing to the cache costs more than a normal input token, 1.25x for the five-minute cache and 2x for the one-hour one, so caching pays for itself after a single read on the five-minute tier. The exceptions run cheaper still: on Claude Opus 5.5 a cache read costs 5% of the base input price ($0.20 per million tokens), and on Claude Fable 5.1 (and the limited-availability Mythos 5.1) it costs 2.5%. Sonnet 5.5 is 5% too: $0.10 per million tokens, per the pricing docs (Anthropic’s launch page now agrees). The Batch API is exactly 50% off both input and output for asynchronous work. The two discounts stack, per Anthropic’s own docs, so a batched job hitting a warm cache is paying a fraction of list. (We get into cross-provider cost-cutting in our free AI APIs for developers roundup.)
A subscription instead of the API. If your use case is interactive — chatting with Claude, coding with Claude Code — you probably don’t want the API at all. Claude Pro is $17/month billed annually or $20/month billed monthly; Claude Max is $100/month for 5x Pro’s usage every five hours and every week. Above it sits Claude Ultra at $200/month, which is what used to be sold as Max 20x, renamed at the same price (Anthropic’s Ultra support article says existing subscribers moved over “with no price change”). Anthropic lists Ultra at 10x Pro’s usage every week. The headline perk is no five-hour limit, but read the footnote: that’s for verified accounts. Anthropic verifies automatically from your account and payment history, and until yours is verified “a temporary limit applies.” So a brand-new Ultra signup shouldn’t count on uncapped sessions on day one. Claude Code runs on that subscription (or on an API key, your call). For a solo dev doing day-to-day work, the subscription usually beats metering every token. This is genuinely the answer for most people asking “is it free”: it isn’t, but it’s flat, which feels close enough.
Max, Ultra and Team now include API credits. For ongoing use, this is the closest thing to free API access Anthropic offers. Per the support article Monthly API credits for Max, Ultra, and Team plans, Max comes with $100 of API credit a month and Ultra with $200. Team gets $20 per Standard seat and $100 per Premium seat, pooled and capped at $500. Pro gets none. The credit covers the Messages and Batches API, the Console Playground, Managed Agents and the Agent SDK, “separate from your plan limits”. It doesn’t cover interactive Claude Code, it doesn’t apply to Claude on Bedrock, Vertex AI or Foundry, and unused credit expires at the end of each billing cycle.
Two catches. First, the credit isn’t automatic. You claim it on claude.ai by linking a Claude Console organization to your plan, and new subscribers have to wait seven days on an eligible plan before they can claim. Second, how you authenticate decides who pays. Run claude -p or the Agent SDK with an API key from that linked Console organization and the credit covers it. Signed in with your plan instead, those runs “draw from your plan’s usage limits and don’t use your API credits.” Anthropic’s Agent SDK article now opens with an October 7 note pointing to these credits.
Claude through a cloud marketplace. You can call Claude via AWS Bedrock, Google Cloud, and similar. Still metered usage — but if you’re a new cloud customer, your provider’s signup credits may apply, which is effectively free Claude until those run out.
Student and startup credit programs. Where they exist, accelerator and education programs sometimes bundle Anthropic credits. Worth checking if you qualify. Don’t build a business plan on it.
What you won’t find is a “use it free up to N requests a day, forever” tier. That’s not how Anthropic prices the API.
How the pricing works
Three things decide what a call costs.
First, which model you pick. Here are the per-million-token prices, checked against Anthropic’s pricing docs on 31 October 2026. Rows marked legacy are the ones Anthropic’s models overview files under “Legacy models (still available)”; they still work and still bill at these prices.
| Model | Input / MTok | Output / MTok |
|---|---|---|
| Claude Haiku 5.5 (prompts up to 100K tokens) | $0.10 | $0.50 |
| Claude Haiku 5.5 (prompts over 100K tokens) | $0.50 | $2.50 |
| Claude Sonnet 5.5 | $2 | $10 |
| Claude Opus 5.5 | $4 | $20 |
| Claude Fable 5.1 | $10 | $50 |
| Claude Haiku 4.5 (legacy) | $1 | $5 |
| Claude Sonnet 5 (legacy) | $2 | $10 |
| Claude Sonnet 4.6 (legacy) | $3 | $15 |
| Claude Opus 5 (legacy) | $5 | $25 |
| Claude Fable 5 (legacy) | $10 | $50 |
The current Sonnet costs $2/$10, and the threatened rise to $3/$15 never happened. Sonnet 5 launched at $2/$10 as introductory pricing due to end on 31 August 2026, with the increase scheduled for 1 September. Anthropic’s pricing docs now state that the $2/$10 rate “is now the standard price” and that the increase “will not occur”. Size a Sonnet budget off $2/$10.
That price carried straight over to Claude Sonnet 5.5, which launched on 28 September at the same $2/$10 and has replaced Sonnet 5 in Anthropic’s “Compare models” table. Sonnet 5 now sits under “Legacy models (still available)”, though the deprecations page still lists it as Active with no retirement before 30 June 2027, so nothing breaks if you leave it alone. Anthropic’s launch page claims Sonnet 5.5 “costs up to 30% less for most work” than Sonnet 5 because it “typically needs far fewer tokens to do the same work”. Same list price, so any saving shows up in token counts, not the rate card. I’d test it on your own prompts before believing the 30%.
The other recent row is Opus. Claude Opus 5.5 arrived on 22 September at $4/$20, which makes the current Opus cheaper than the one it replaces, and Anthropic’s models overview now says: “If you’re unsure which model to use, start with Claude Opus 5.5 for most workloads.” If you’re still calling Opus 5 at $5/$25, moving up is also moving down in price.
And Opus is no longer the top of the ladder. Claude Fable 5.1 sits above it at $10/$50, two and a half times Opus 5.5. Fable 5 has the same base price (though cache reads cost more on it), but it’s also on the legacy list, so Fable 5.1 is the top rung to plan around. The May version of this article described a three-rung ladder — Haiku, Sonnet, Opus — and the current lineup now has a fourth rung above Opus. If you’ve been assuming Opus is your worst case on cost, it isn’t.
At the bottom, Claude Haiku 5.5 has taken over as the current Haiku, and it’s the first model on this list whose price depends on how long your prompt is. Up to 100,000 tokens it’s $0.10/$0.50. Go over that and the whole request bills at $0.50/$2.50, five times more, output included. Cache reads and writes count toward the 100,000, so a cached history doesn’t keep you under it. A chatbot that keeps re-sending its whole history will cross that line without anyone deciding to, so budget for the higher band if your prompts grow. Haiku 4.5 still runs at $1/$5, and its status depends on which Anthropic page you read: the models overview files it under legacy, while the deprecations page lists claude-haiku-4-5-20251001 as Active with no retirement before 15 October 2026. That’s a floor, not a date: Anthropic promises “at least 60 days’ notice before model retirement”, and no deprecation notice for Haiku 4.5 has been published. As of 31 October the 15 October floor has passed and it’s still listed Active, so the earliest it can now retire is 60 days after a notice that hasn’t appeared yet. One more thing before you compare the two Haikus on price alone: Haiku 5.5 “uses the same newer tokenizer as Claude 4.7 and later models, so the same text counts as approximately 30 percent more tokens than on Claude Haiku 4.5.”
The gap between cheapest and priciest is now a hundred times on both, comparing Fable 5.1 against Haiku 5.5’s up-to-100K band. Against Haiku 5.5’s over-100K band it’s twenty times on both. Earlier versions of this page said ten times, which was the gap from Haiku 4.5 before Haiku 5.5 came along. Picking the smallest model that actually does your job is the single biggest budget decision you’ll make.
Second, input versus output tokens. You’re billed separately for what you send (input) and what Claude generates (output), and output tokens cost more: five times the input rate on every model in the table above. An app that sends a huge prompt and gets a one-line answer back is cheap. An app that sends a small prompt and gets a 2,000-word essay is not.
Third, how much context you send, which is the one that sneaks up on people. Every API call is stateless — there’s no server-side memory — so if you’re building a chatbot or an agent, you re-send the conversation history (and any retrieved docs, and the system prompt) on every single turn. A 20-turn conversation means turn 20 is paying for ~19 turns of history as input. Agentic coding tools are the extreme version: they shovel file contents, tool outputs, and long instructions into context, which is exactly why a serious coding session costs real money. Context size is the driver. Get a handle on it and the bill drops.
Rough intuition: one short Q&A on a small model is a fraction of a cent. A typical Sonnet request with moderate context is in the cents. A long agentic session with lots of files in context and long generated diffs can be a few dollars. Your actual numbers depend entirely on your workload — the only way to really know is to run a realistic week and look at the dashboard.
Rate limits, explained
Anthropic uses usage tiers. As of October 2026 they’re named Evaluation, Start, Build, Scale and Custom — if you’ve read an older guide referring to “Tier 1” through “Tier 4,” that naming is gone. New organisations may start in Evaluation, whose limits Anthropic doesn’t publish (“limits below the standard limits shown on this page”). Start, Build and Scale are documented; Custom is by arrangement with sales. For Claude Sonnet 5.5 and Sonnet 5, per the rate-limits docs (RPM is requests per minute, TPM tokens per minute). Sonnet 5.5 and Sonnet 5 each get their own bucket. Token limits are the same for both, but Sonnet 5.5 now gets a higher request ceiling at every tier:
| Tier | RPM (Sonnet 5.5) | RPM (Sonnet 5) | Input TPM | Output TPM | Spend cap/mo |
|---|---|---|---|---|---|
| Start | 4,000 | 1,000 | 2M | 400K | $500 |
| Build | 10,000 | 5,000 | 5M | 1M | $1,000 |
| Scale | 13,000 | 10,000 | 10M | 2M | $200,000 |
Until late October the two Sonnets had identical numbers. The RPM bump for Sonnet 5.5 showed up between our 24 October and 31 October checks; token-per-minute limits and spend caps didn’t move. If you’re still pinned to Sonnet 5 and hitting request limits, upgrading gets you up to four times the RPM at the Start tier for the same $2/$10. The same bump applies to Opus 5.5 and Haiku 5.5 (4,000/10,000/13,000 RPM across Start/Build/Scale), while Opus 5 and Haiku 4.5 stay at 1,000/5,000/10,000.
Note the last column, because it surprises people: the Start tier caps you at $500 of API spend per month whether you want the ceiling or not. Limits are also set per model class, not one blanket number for your org. Claude Fable 5.1 and Fable 5 run lower token limits at the Start tier (4,000 RPM but only 500,000 input tokens/min), and that is one combined limit across both models. Splitting traffic between them doesn’t double it.
One genuinely useful detail that isn’t widely known: cached input doesn’t count toward your input-tokens-per-minute limit on most models. Anthropic’s own worked example is a 2,000,000 ITPM limit with an 80% cache hit rate effectively processing 10,000,000 input tokens per minute. If you’re rate-limited rather than cost-limited, prompt caching is the fix for both.
You can request increases through the Console if you’ve got a legitimate need outgrowing the automatic tiering.
What that means in practice: don’t load-test from a fresh account, because you’ll hit the per-minute caps immediately and conclude the API is “rate-limited” when really you’re just on the lowest tier. Build retry-with-backoff from day one — when you do hit a limit, the API tells you, and a well-behaved client backs off rather than hammering. And plan your launch around tiering: if you expect a spike, get your account spending (and aging) ahead of time, or request a limit increase before launch, not the morning of.
Exact rate-limit numbers per tier change, so check the current docs rather than trusting a number you read in a blog post. (Yes, including this one — which is why the date at the top of this page is the date every number on it was last checked.)
Models retire, too, and faster than most people plan for. Next on the calendar is Claude Sonnet 4.5 (claude-sonnet-4-5-20250929): deprecated on 30 September 2026, retiring on 30 November 2026, with claude-sonnet-5-5 named as the replacement. Claude Opus 4.1 was retired on 5 August 2026. Opus 4 and Sonnet 4 went on 15 June 2026; Sonnet 3.7 and Haiku 3.5 on 19 February 2026. Anthropic commits to at least 60 days’ notice, and requests to a retired model ID fail outright on the first-party API even where the same model lingers on Bedrock or Google Cloud. If you hardcoded a dated model string a year ago, go and check it.
How to use Claude without burning money
A few habits do most of the work. Turn on prompt caching if any meaningful chunk of your prompt repeats across calls; it’s the big, easy win. Run non-urgent work through the Batch API for roughly half off: classification jobs, bulk summarization, overnight processing.
Pick the smallest model that works. Haiku-class for extraction, classification, routing; Sonnet for most coding and reasoning; Opus when the task needs it (Anthropic’s own suggested starting point is now Opus 5.5, but that’s a quality default, not a budget one). Test down the ladder, not just up.
Trim your context. Don’t re-send the entire conversation when a summary will do, and don’t dump whole files when a function will do.
Set a monthly spend cap in the Console so a runaway loop or a bad deploy can’t drain the account. Watch the usage dashboard, especially after you ship something new — costs that surprise you in week one are cheap to fix; costs that surprise you in month three aren’t.
And for interactive use, lean on a subscription: if you’re personally coding with Claude all day, Claude Code on Pro, Max or Ultra is almost certainly cheaper than the API. Our Claude Code tips post covers getting the most out of it.
API or subscription?
Quick way to decide. If you’re building an app or automation that calls Claude programmatically, you need the API. A subscription only helps here through the monthly API credit that comes with Max, Ultra and Team (above), and you spend it with an API key from your linked Console organization. Past that, you’ll pay per token. Optimize with caching, batching, and model choice.
If you’re personally coding, chatting, or doing day-to-day work with Claude, get a subscription (Claude Pro, Max or Ultra, with Claude Code). Flat fee, big allotment, no per-token anxiety. That’s what most “is the API free” askers actually want.
Just prototyping? Start with the new-account trial credit on the API, or the free tier of the Claude consumer app, then decide.
And if you’re already deep in AWS or Google Cloud, consider calling Claude through that provider so it rolls into your existing cloud bill, possibly against your cloud credits.
TL;DR
The Claude API isn’t free on an ongoing basis. New accounts get a small one-time trial credit; after that it’s pay-as-you-go — prepaid credits, billed per million tokens. Cost comes down to model choice times (input plus output tokens) times how much context you send; output costs more than input, and context size is the main driver. Rate limits scale by usage tier — new accounts start low and get promoted automatically as they age and spend — so build backoff/retry and don’t load-test from a fresh account. The cheapest way to use it: prompt caching, the Batch API, the smallest model that works, trimmed context, a spend cap, and an eye on the usage dashboard. If you want flat-rate Claude for your own interactive work, that’s a subscription (Claude Pro, Max or Ultra, with Claude Code). Max, Ultra and Team also carry a monthly API credit ($100 on Max, $200 on Ultra, $20 or $100 per Team seat) that you claim by linking a Claude Console organization to your plan; Pro doesn’t get one.
Pricing and limits change. Confirm the current numbers on Anthropic’s official pages before you budget.
Update notes: the 31 October 2026 check found Anthropic had renamed Max 20x to Claude Ultra at the same $200 price, listed at 10x Pro’s usage every week, with no five-hour limit for verified accounts (unverified accounts get a temporary limit), leaving Claude Max as the single $100 plan at 5x Pro; it also found that Sonnet 5.5 now has higher RPM limits than Sonnet 5 (earlier versions said the two were identical). The 19 October 2026 check settled Sonnet 5.5’s cache-read price at $0.10 per million tokens (earlier versions said to budget $0.20) and noted that Haiku 4.5 passed its 15 October retirement floor with no deprecation notice. The 10 October update added Claude Haiku 5.5 and its prompt-length pricing, corrected the cheapest-to-priciest gap (100x, not 10x), and rewrote the Agent SDK billing note around the Max and Team API credits.
Want the wider view — which AI APIs actually have usable free tiers, and which “free” offers come with a catch? Read Best Free AI APIs for Developers (2026) — With Real Rate Limits. For non-AI freebies, Every Free Credit and Promo Code for Developer Tools is the companion list. More in the AI Tools section.
Frequently Asked Questions
Does the Claude API have a free tier?
Not an ongoing one. New Claude Console accounts usually get a small amount of free trial credit to get started, but after that it's pay-as-you-go — you prepay credits and get billed per million input and output tokens. There's no permanent free quota.
How much does the Claude API cost?
Pricing is per million tokens and depends on the model. As of 31 October 2026: Claude Haiku 5.5 is $0.10 input / $0.50 output per million tokens for prompts up to 100,000 tokens and $0.50/$2.50 above that, Claude Sonnet 5.5 is $2/$10, Claude Opus 5.5 is $4/$20, and Claude Fable 5.1 is $10/$50. Older models still work at their own prices: Haiku 4.5 ($1/$5), Sonnet 5 ($2/$10, the introductory rate Anthropic made the standard price), Sonnet 4.6 ($3/$15), Opus 5 ($5/$25) and Fable 5 ($10/$50). Anthropic's models overview lists them as legacy, while its deprecations page still shows them as Active. Output costs five times input on each of these models, and context size is the main cost driver.
Is Claude Code free?
Not through the API directly. Claude Code needs either a Claude Pro, Max or Ultra subscription or an Anthropic API key. The subscription includes a usage allotment; the API key is pay-as-you-go. New API accounts get some trial credit, but sustained use means paying.
How do I get free Claude API credits?
Sign up for a Claude Console account to get the new-account trial credit. Beyond that, look for startup or student credit programs if you qualify, or new-customer cloud credits if you reach Claude through AWS Bedrock or Google Cloud. If you're on Max, Ultra or Team, you can also claim a monthly API credit ($100 on Max, $200 on Ultra, $20 or $100 per Team seat) by linking a Claude Console organization to your plan. Pro doesn't get one. None of it is unlimited.
Comments