Twitter (X) Ads Cost 2026: CPM, CPC, Minimums and a $1,000 Test

X does not publish an average price for its ads. You pay in an auction, there is no minimum spend, and the lowest bid X accepts is $0.01 per desired action. The third-party numbers that exist put X CPM (cost per 1,000 impressions) anywhere from $0.64 (Gupta Media’s blended 2026 average) to $2.09 (Hootsuite’s own 2025 campaigns) to $25 to $45 for narrow developer audiences. Below is every dated benchmark worth using, what X actually bills you for, what Premium and Premium Business cost with their ad credits, a worked $1,000 test, current ad specs, and an honest answer on which niches have an audience on X.

How much do Twitter (X) ads cost? CPM per 1,000 impressions, CPC and CPE

Every number in this table comes from a vendor, agency or tracker, not from X. The SpaceX S-1 analysis by Digital Applied confirms the filing gives no platform CPMs, and X’s own help center says the price per action “is not a fixed rate”. Read each row with its source and audience in mind.

Why the X CPM numbers disagree by up to 70x

The cheap numbers are blended averages across every country and objective. X’s programmatic native inventory averaged a $0.39 CPM as of December 2024, and Japan, not the US, supplies about 29% of X’s weekly programmatic impressions versus 21% for the US, per data cited in the same PPC Land report. A US campaign aimed at a professional audience competes with far fewer cheap impressions. That is why the developer-targeting figure is 40 to 70 times the blended one. Plan with your own targeting, then use X’s bid guidance in Ads Manager, which shows what other advertisers bid once you enter a budget and dates.

Timing moves the price too. The SpaceX S-1 states that X’s advertising spend tends to be highest in the fourth quarter and lowest in the first quarter. If you test in November, expect to pay more for the same reach than you would in February.

What X actually charges you for: objectives, billing events and bid types

Your cost depends first on what you are billed for. X’s Ads API documentation maps each objective to a pricing model. Ads Manager labels can differ slightly, and some objectives offer a “Pay by” choice: for Video views, X lets you pick “Pay by impressions (recommended)” or “Pay by Video views”.

X offers four bid types:

1. Automatic bid. X picks the bid to get the most results within your budget. X recommends it for new advertisers and says it is the most likely to spend your full daily budget.
2. Maximum bid. You set the most you will pay per action. Ads Manager pre-fills the midpoint of its bid guidance until you change it. The API notes this is not available for Reach or Followers.
3. Target bid. Available for Website traffic (link clicks), Reach, App installs and App re-engagements. X optimizes toward a daily average at or below your target, and the API reference says it aims for daily averages within 20% of the bid.
4. Target cost per action. For the Sales objective (target cost per purchase, per download). X says it needs about 10 conversions before the algorithm paces toward your target, and that a target set too low limits spend.

The auction is second price with a quality adjustment. X’s own example: bid $1.40, beat an otherwise identical ad bidding $1.00, and you are charged $1.01. Ads with better “resonance, relevance and recency” win cheaper, so weak creative raises your real CPM even with the same bid.

X Ads minimums, budgets and billing

• Minimum spend: none. X states “There is no minimum spend required for X Ads”. Minimum bid is $0.01 per action.
• Budget setup: in the default structure you must set a daily budget and a start date for each ad group, with an optional spend cap. You cannot choose the hours of day your ads serve.
• Pacing: standard delivery is the default. Turning it off switches to accelerated delivery, which can exhaust the daily budget early in the day.
• Payment: a credit or debit card is required to launch; advertisers spending over $5,000 a month can request an insertion order, per Hootsuite’s guide.
• Final billing: X’s API docs say billing data may not be final until three days after the event, and the final billable number is corrected for spam and low-quality traffic, so it will be lower than the early figure.
• New-advertiser credit: as of this writing, X’s advertising page offers “$500 in ad credit when you spend $500” for eligible new advertisers, terms apply.

X Premium and Premium Business prices, and the ad credits they include

Personal X Premium tiers do not come with ad credits. The business tiers do, for a limited period, and the credit can equal the monthly fee.

The fine print matters. Per X’s terms: credits only apply if you pay for the subscription through Stripe on a tier that includes them; monthly credits must be activated and used within 30 days, annual ones within 90 days; unused credits vanish if you cancel, upgrade or downgrade; and X can change or revoke the offer at any time. X’s Premium Business comparison page also lists a gold checkmark and Premium+ and SuperGrok access on every tier, and X’s advertising page lists “Mentions Boost” for Premium Business users to boost organic posts that mention their brand.

The practical read: if you would spend at least $200 a month on X ads anyway, Premium Business Basic is close to free while the credit offer lasts. If you would not, you are paying $200 a month for a checkmark. Keep in mind the European Commission fined X €120 million in December 2025, finding the paid blue checkmark deceptive because anyone can buy one without X confirming who is behind the account. The badge buys features, not trust.

A worked $1,000 X ads test for a service business

This plan uses the Website traffic objective, because it bills per link click and lets you read cost per lead directly. The click costs are the published benchmarks above. The landing page conversion rate and client economics are assumptions you should replace with your own.

Setup: $1,000 over 20 days at $50 a day, automatic bid, standard delivery, three ad groups:

What $1,000 buys at three click costs:

Your break-even cost per lead: say a new client is worth $6,000 in first-year fees, you close 20% of leads, and you will spend up to 20% of first-year revenue to win a client. Maximum cost per client is $6,000 × 20% = $1,200. Maximum cost per lead is $1,200 × 20% = $240. In that case even the stress-test scenario passes on paper. Whether the leads close is the real test, so track every X lead to a sales outcome with UTM tags and a CRM source field.

Pass/fail rules for the test (my rules, not X’s):

1. Day 3: if an ad group has spent $60 with zero link clicks, its creative or targeting is broken. Pause it and move the budget.
2. Day 10: compare cost per lead against the same number from your current best channel. Above 2x and no qualified conversations: cut it.
3. Day 20: pass only if at least one lead reached a sales conversation and the projected cost per client is under your break-even. Clicks and engagements alone do not count.
4. Do not judge billing until three days after the test ends, because X finalizes spam-corrected billable numbers on that lag.

If you qualify for X’s spend-$500-get-$500 new-advertiser credit, the same delivery can cost less cash, but read the terms first. Benchmarks for other channels are in our cost per lead by industry breakdown.

Best niches to grow fast on X in 2026: what the evidence supports

No public dataset ranks “fastest-growing niches” on X. The lists ranking higher for this query come mostly from growth-tool vendors and show no method. What does exist is research on why people use X, what they read there, and where advertisers spend. Together it points to the same topics.

Read together, the niches with demonstrated audiences on X are: news and current-events commentary, politics and policy, sports, technology and software (including AI), markets and financial services, gaming, and entertainment. Because most users read rather than post, an account that consistently posts useful takes in one of these topics faces less competition for attention than the user counts suggest.

For a professional-service firm, the move is a narrow slice of one of those topics where your expertise is the news: tax law changes for a CPA firm, deal-market commentary for an M&A advisor, AI tooling for an IT consultancy. Score your niche before you commit:

A total of 8 or more out of 10 is worth a 90-day organic test before you spend on ads. Under 5, put the effort into LinkedIn or search. For more creative patterns that work on X and other networks, see these social media ads examples.

Who is on X in 2026, and is it worth it for B2B versus LinkedIn?

Pew’s 2026 survey of 5,511 US adults found 18% use X, down from 21% in 2025, while 31% use LinkedIn. The gap is widest among the people most B2B firms sell to:

Source: Pew Research Center, survey Jan. 30 to June 18, 2026. Globally, the SpaceX S-1 reports about 550 million monthly active users across X and Grok combined as of March 31, 2026, a figure that is not comparable with Twitter’s old daily-user metric. For more channel data, see our social media marketing statistics for 2026.

On cost, X is cheaper to enter. LinkedIn requires a $10 minimum daily budget and a $100 minimum lifetime budget; X has no minimum. In the one benchmark that compares the two on the same audience (developers, 2026), X CPMs ran $25 to $45 against LinkedIn’s $60 to $120, per daily.dev. What you lose is LinkedIn’s job-title and company targeting.

I treat X as a second or third channel for most service firms, run next to a working LinkedIn ads program rather than instead of it. If you want a second opinion on where your next ad dollar should go, you can book a consultation.

X ad specs for 2026: character limits, image and video sizes

From X’s creative ad specifications, with card image rules from the Ads API creatives docs. X notes that not every format is available to self-serve advertisers.

For a professional-service ad, the cheapest win is the copy limit: put the specific number or claim in the first line, because 257 characters with a link is the whole pitch. If you plan retargeting across platforms, our retargeting playbook covers the audience setup.

Brand safety and the advertiser picture in 2026

Before you put a firm’s name next to X content, know what has changed and what is still uncertain.

• Revenue fell, then partly recovered. The SpaceX S-1 reports X’s advertising revenue fell $595 million in 2024 “due to the loss of advertising partners,” rose $116 million in 2025, then fell $100 million year over year in Q1 2026 during “an overhaul of the Company’s advertising platform.”
• Forecasts are modest. EMARKETER estimates (labeled forecasts, not reported revenue) put X’s global ad revenue at about $2.26 billion in 2025 and $2.46 billion in 2026, versus the $4.51 billion Twitter reported for 2021, per Digital Applied.
• Big brands are back, by X’s account. X’s ad chief said at CES in January 2026 that 97 of the top 100 advertisers have returned. Sensor Tower estimates show large year-over-year spend increases from Comcast, Google, Dell and American Express, per Digiday. X declined to share its own spend figures.
• Independent measurement has gaps. Integral Ad Science’s pre-bid filtering on X covers only Vertical Video in the US. Post-bid measurement covers the timeline in six markets and, since January 2026, profiles. Search, replies and other surfaces sit outside IAS post-bid scope, per PPC Land’s review of IAS’s partnership document (March 2026).
• Regulatory and AI risk. Beyond the EU’s €120 million DSA fine, which also cited X’s ad repository and researcher data access, Digiday reported that Grok generated sexualized images of women and minors, prompting investigations in the UK and Europe. X’s own brand suitability playbook says Grok now screens content in its pre-bid and adjacency controls, per PPC Land.

For a regulated or reputation-sensitive firm (law, wealth management, healthcare), my default is: use keyword and topic exclusions, run on the timeline rather than profiles or replies where you can, and review placements weekly. If your compliance team cannot sign off on that residual risk, skip paid X and keep an organic presence only.

Frequently asked questions

How much do Twitter ads cost per 1,000 impressions?

X does not publish an average CPM. Third-party figures range from about $0.64 (Gupta Media's blended 2026 average, via Digiday) to $2.09 (Hootsuite's own 2025 campaigns) to $25 to $45 for narrow developer audiences (idlen.io, 2026). Blended averages include cheap non-US inventory, so a US B2B campaign will usually pay well above the low end.

Is there a minimum budget for X ads?

No. X's help center states there is no minimum spend for X Ads, and the lowest bid it accepts is $0.01 per desired action. You do need a card on file and, in the default setup, a daily budget and start date for each ad group. LinkedIn, by comparison, requires $10 a day and $100 lifetime per campaign.

How much does it cost to advertise on Twitter per day?

Whatever daily budget you set. X bills per action in an auction, so $20 a day buys about 27 link clicks at Hootsuite's 2025 average CPC of $0.74, or about 7 at a $3.00 CPC. Start around $50 a day split across two or three ad groups so each gets enough data within two to three weeks.

Do X Premium subscribers get ad credits?

Personal Premium tiers ($3, $8 and $40 a month) list no ad credits. Premium Business Basic ($200 a month) gets $200 in ad credit every 30 days, and Full Access ($1,000 a month) gets $1,000, for a limited period per X's terms updated April 1, 2026. Credits must be used within 30 days and are void if you cancel.

What are the best niches to grow fast on X in 2026?

No public dataset ranks niches by growth. Pew research shows X users come for entertainment, politics, news, sports and pop culture, and 57% regularly get news there. Sensor Tower estimates show the most ad spend in media, shopping, software, financial services and gaming. Pick a narrow slice of one of those where your expertise makes news.

How many people use X compared with Threads?

In Pew's 2025 survey, 21% of US adults used X and 8% used Threads. Pew's 2026 survey put X at 18%. Globally, the SpaceX S-1 reports about 550 million monthly active users across X and Grok combined as of March 31, 2026, a combined figure that is not directly comparable with standalone app counts.

Are X ads better than LinkedIn ads for B2B?

Usually not as a first channel. Pew's 2026 data shows 51% of US college graduates use LinkedIn versus 22% for X, and LinkedIn offers job-title targeting. X can work as a cheaper second channel when your buyers follow news, markets, policy or tech in real time, and when you can upload a CRM list for warm targeting.