Brazil 2026 Election Odds: The AI Model Verdict Vs. The Kalshi Market

Kalshi's market on who wins Brazil's 2026 presidential election has a clean shape at the top and a messy one underneath: an 80-year-old incumbent is the runaway favorite, his leading challenger is a senator running because his father is barred from the ballot, and the country's most talked-about right-wing governor is not even in the race. We asked six AI models to score the entire 13-candidate field without letting any of them see a single price, working only from a fetched, dated summary of the polls and the eligibility rulings. They came back agreeing with the market on the favorite, slightly more confident in him than the crowd once you strip out the spread, and much less impressed than the market by one newcomer near the bottom.

The Quick Answer

Kalshi's market makes Luiz Inácio Lula da Silva the clear favorite at 67 cents to win the 2026 Brazilian presidential election. Our price-blind, six-model panel lands slightly higher, blending the incumbent to 68.0%, with Flávio Bolsonaro, the Liberal Party standard-bearer, a distant second at 23.8% (market 26¢). Once you re-scale the market for its overround, the panel is a touch more bullish on Lula than the crowd, and its one real fade is on Free Brazil Movement founder Renan Santos, priced at 8¢ but blended to just 1.5%.

• Market Favorite: Luiz Inácio Lula da Silva, 67¢ (about 63% once re-scaled for the spread across the board)
• Model Favorite: Luiz Inácio Lula da Silva, 68.0% blend
• Distant Second (Both): Flávio Bolsonaro, 23.8% blend vs a 26¢ price
• Where The Panel Fades The Market: Renan Santos, 1.5% blend vs an 8¢ price
• The Panel's Logic In One Line: the incumbent leads every fetched July poll in the first round and wins every tested runoff scenario, while the strongest opposition names are legally eliminated and the rest of the field is a rounding error.

The two-round settlement mechanic, the eligibility rulings that reshaped the field, the full model-by-model numbers, the dark horses, and a plain-English guide to reading these prices are all below.

Want the panel's verdict when this market settles? We grade every price-blind call in public once a market resolves. Grab the re-scored board and the settled result through the free email capture at the end of this article; no promo code, no pick sold.

New to Kalshi's event contracts? Start with our guide to trading events on Kalshi before reading the board below.

The Market At A Glance

Venue	Kalshi, a CFTC-regulated event-contract exchange (18+; availability varies by state, as of August 2026)
The Event	"Brazil Presidential election winner?" (Kalshi event KXBRPRES-26), 13 separate yes/no contracts, one per named candidate
The Contract	Each market resolves YES only if that candidate wins the 2026 Brazilian presidential election, per Kalshi's settlement rules
The Key Rule	Brazil uses a two-round majority system: a first round on October 4, 2026, and a runoff on October 25, 2026 if no candidate wins more than 50% of the valid first-round vote. The board settles on the overall winner, so a first-round leader can still lose the presidency in the runoff.
Prices Below As Of	The morning of August 3, 2026 (Kalshi last-trade and bid/ask midpoints from candlestick history; several contracts are thin). Prices on this page refresh during the campaign; the timestamp here carries the reading these numbers reflect.

These are model estimates, not predictions of fact and not financial advice. Kalshi event contracts trade on a CFTC-regulated exchange; you must be 18 or older and in an eligible state to participate. This piece takes no political position; every fact below is sourced and dated, and all candidates and claims are described neutrally.

Why The Field Looks The Way It Does

Before any probability makes sense, the field itself needs explaining, because three of the names a casual observer would expect to see near the top are gone for legal or strategic reasons. This is where a reader can actually get better-informed, and it is all fetched and dated.

• Jair Bolsonaro Cannot Run. Brazil's Superior Electoral Court ruled the former president ineligible until 2030 in a 5-2 decision on June 30, 2023, for abuse of power, and as of 2026 he is under house arrest and convicted in the 2022 coup case. His 1.1¢ contract is a legal near-impossibility, not a live candidacy.
• Eduardo Bolsonaro Cannot Run Either. The former president's son was ruled ineligible for 12 years on June 16, 2026, tied to a coercion conviction in the coup case.
• Tarcísio De Freitas Chose Out. The São Paulo governor, often floated as the right's strongest option, is not legally barred, but he opted to seek re-election as governor. Under Brazilian law an executive must resign months ahead to run for a different office, which rules him out of the presidential race.
• What Is Left On The Right Is Flávio Bolsonaro, the senator son whom Jair endorsed to carry the Liberal Party (PL) flag, plus regional governors and a movement newcomer. The consolidated opposition standard-bearer is therefore a first-time presidential candidate who trails the incumbent and, per reporting, faces his own court risk.

That reshaped field is the backdrop for every number the panel produced. The models did not have to guess at any of it; the eligibility facts were handed to them in a dated card, which is why even the seats most cautious about Lula still priced the barred and non-running names near zero.

Where The Polls Stand

The panel worked from two fetched July 2026 first-round polls, both neutral voting-intention surveys:

• Quaest, fieldwork July 10 to 13, 2026 (n=2,004): Lula 39%, Flávio Bolsonaro 33% in the first round.
• Datafolha, fieldwork July 22 to 23, 2026 (n=2,004): Lula 40%, Flávio Bolsonaro 32%, Ronaldo Caiado 4%, Romeu Zema 3%, Renan Santos 3%.

Both show Lula ahead by 7 to 8 points in the first round and winning the tested runoff scenarios, consistent with an earlier BTG/Nexus poll from December 2025 that had Lula 42% and Flávio 34%. The one number every model kept returning to is the runoff: because the board pays on the overall winner, and because Lula wins the head-to-head polls after the anti-incumbent vote consolidates behind Flávio, the two-candidate math points at the incumbent directly rather than by inference from a split first round.

Here is the full field, sorted by their market price, with each model's price-blind number and the re-scaled market price.

The Board

Outcome	Role / status	Market	AI blend
Luiz Inácio Lula Da Silva	Incumbent president (PT), age 80, seeking a 4th term	67¢	68.0%
Flávio Bolsonaro	Senator, Liberal Party (PL) pre-candidate	26¢	23.8%
Renan Santos	Free Brazil Movement founder	8¢	1.5%
Ronaldo Caiado	Governor of Goiás	2.4¢	1.9%
Romeu Zema	Governor of Minas Gerais	0.5¢	1.1%
Field / Someone Not Listed	—	—	~3.5%

Every other contract on the board blended to essentially zero: Jair Bolsonaro, Eduardo Bolsonaro, and Tarcísio de Freitas for the eligibility and non-candidacy reasons above, and Michelle Bolsonaro, Fernando Haddad, Ciro Gomes, Aldo Rebelo, and Ratinho Júnior for negligible polling support. Their full numbers are in the table below.

Every Seat's Number

Every model scored all 13 contracts price-blind. The blend is the simple mean of the six numbers in each row.

Outcome	Market (implied)	Fable	Opus	Sonnet	GLM	Kimi	DeepSeek	Blend
Luiz Inácio Lula Da Silva	67¢ (62.9%)	64.0	67.0	64.9	67.0	67.0	78.0	68.0%
Flávio Bolsonaro	26¢ (24.4%)	24.0	24.0	24.7	24.0	26.0	20.0	23.8%
Renan Santos	8¢ (7.5%)	2.0	2.0	2.1	2.0	1.0	0.0	1.5%
Ronaldo Caiado	2.4¢ (2.3%)	3.0	1.5	3.1	2.0	2.0	0.0	1.9%
Romeu Zema	0.5¢ (0.5%)	2.0	1.5	2.1	0.0	1.0	0.0	1.1%
Ratinho Júnior	0.1¢ (0.1%)	0.0	0.0	0.0	1.0	0.0	0.0	0.2%
Jair Bolsonaro	1.1¢ (1.0%)	0.0	0.0	0.0	0.0	0.0	0.0	0.0%
Ciro Gomes	0.5¢ (0.5%)	0.0	0.0	0.0	0.0	0.0	0.0	0.0%
Michelle Bolsonaro	0.4¢ (0.4%)	0.0	0.0	0.0	0.0	0.0	0.0	0.0%
Fernando Haddad	0.2¢ (0.2%)	0.0	0.0	0.0	0.0	0.0	0.0	0.0%
Eduardo Bolsonaro	0.1¢ (0.1%)	0.0	0.0	0.0	0.0	0.0	0.0	0.0%
Aldo Rebelo	0.1¢ (0.1%)	0.0	0.0	0.0	0.0	0.0	0.0	0.0%
Tarcísio De Freitas	0.1¢ (0.1%)	0.0	0.0	0.0	0.0	0.0	0.0	0.0%
Field / Someone Not Listed	—	5.0	4.0	3.1	4.0	3.0	2.0	3.5%

These are model estimates, not predictions of fact and not financial advice. Market figures are Kalshi last-trade and bid/ask midpoints from candlestick history; the "implied" percentage re-scales each price against the board's total, which sums to about 106.5 cents because each of the 13 contracts trades with its own spread. Every number in this piece gets graded in public once the market settles; you can check the method against past calls on the full graded scoreboard.

A note on the panel. Our standard Model Verdict lineup runs eight seats. On this run two were unavailable: the ChatGPT (codex) and Gemini (agy) seats were not installed on the machine that produced these verdicts. This board therefore runs on six models: Claude Fable, Claude Opus, Claude Sonnet, GLM, Kimi, and DeepSeek. All six scored the field on August 3, 2026 without seeing any prices, working from a fetched, dated summary of the 2026 polls and eligibility rulings. After the first round, each model then read the other seats' anonymized numbers and reasoning and could revise; the numbers above are those revised, round-two estimates. Every model's number also lands in our back-tested Model Verdict ledger, where past price-blind calls are graded against settled results so the method is measured, not just asserted.

Luiz Inácio Lula Da Silva: The Incumbent The Panel Likes Slightly More Than The Market

The panel's top call is also the market's, and the reasoning was consistent across all six seats: Lula is the incumbent, he leads every fetched July poll in the first round by 7 to 8 points, and he wins every tested runoff scenario against a fragmented, court-constrained opposition. Sonnet summed it up in one line, calling him "the clear favorite, incumbent, leads every fetched first-round poll by 7-8pts, and wins tested runoff scenarios against Flávio Bolsonaro, the only viable opposition standard-bearer after Jair, Eduardo, and Tarcísio were all knocked out by ineligibility or filing rules."

The mild surprise is direction. Lula's 67¢ price looks like it sits right on top of the panel's 68.0%, but the raw prices on this board sum to about 106.5 cents, so once you re-scale for the spread the market's Lula is closer to 63%. The panel is therefore a few points more confident in the incumbent than the crowd, not less. The seats that pulled the blend up were the ones that leaned hardest on the runoff math: as Kimi put it, "the runoff polls, which already reflect consolidation of the anti-Lula vote behind Flávio, still show Lula ahead, so the two-candidate math favors him directly rather than requiring inference from the first round."

The panel is not unanimous on how confident to be, and that spread is honest. DeepSeek sat highest at 78%, calling Lula "the dominant favorite" whom the opposition cannot catch. The more cautious seats, Fable and Sonnet, held near 64%, and their reason is the same one a careful reader should keep in view: Lula is 80 years old, the vote is nearly three months out, and Brazilian runoffs have a history of tightening as the anti-incumbent vote consolidates. None of the models is predicting he loses; they are pricing the roughly one-in-three chance that age, a campaign swing, or runoff consolidation produces a different winner by October 25. That market-versus-blend gap is the single number a reader should weigh most.

Flávio Bolsonaro: The Standard-Bearer By Default

Here is the board's quiet story, and the panel named it cleanly: the consolidated opposition candidate is a senator who is running largely because the two Bolsonaros with higher ceilings are legally out. Flávio Bolsonaro carries his father's endorsement and the Liberal Party's machinery, and he polls a real 32 to 34% in the first round, which is why he is the clear second at 26¢ and a 23.8% blend. This is not a token number; it is the honest probability that the anti-incumbent vote consolidates behind him and a runoff breaks his way.

The most interesting revision on the entire board happened here. Kimi started round one with Flávio at 34%, near his raw first-round poll share, then cut him to 26% after reading the other seats. Its stated reason is a real analytical correction, not conformity: "my 34% on Flávio effectively double-counted runoff-consolidation risk, since the head-to-head runoff polls already show Lula winning after consolidation," and it added that "Flávio's personal court risk lowers his individual odds even where the broader right-wing bloc's chance" is higher. Opus moved the opposite way by a hair, nudging Flávio from 21% up to 24% after peers stressed that Lula's roughly 40% first-round share means a current majority prefers someone else, keeping "a consolidated Flávio runoff" as the live threat. The two arcs meeting near 24% is the panel doing exactly what the convergence round is for.

Where The Panel Fades The Market: Renan Santos

The one place the panel and the market truly disagree is near the bottom of the board. Renan Santos, the Free Brazil Movement founder confirmed as a presidential candidate in July 2026, carries an 8¢ price, third-highest on the board and re-scaled to about 7.5%. The panel wants almost none of it: every seat priced him between 0% and 2.1%, for a 1.5% blend.

The gap is a clean read of the ceiling problem. Santos polled just 3% in the July Datafolha survey, and as a first-time movement candidate with no path through a Lula-versus-Flávio runoff, the models see his realistic win probability as a fraction of his market price. The market may be paying for narrative, a young, online, insurgent candidacy that generates attention, while the panel prices the runoff structure that narrative has to survive. This is the board's sharpest market-versus-model divergence, and it is the number to watch: if Santos climbs in the polls, the panel is wrong and the 8¢ is cheap; if he stalls near 3%, the price is the one the panel is fading.

Dark Horses The Panel Won't Fully Dismiss

Below the top three, a few contracts drew a real number from at least one seat:

• Ronaldo Caiado (2.4¢, Blend 1.9%) is the governor of Goiás and the highest-polling of the regional executives at 4% in Datafolha. Four seats gave him a small real number (Sonnet 3.1%, Fable 3.0%, GLM and Kimi 2.0%), pricing the scenario where a governor consolidates the non-Bolsonaro right. It is the most credible "someone other than Flávio carries the right" bet on the board.
• Romeu Zema (0.5¢, Blend 1.1%) is the governor of Minas Gerais, a large and electorally pivotal state. The panel actually prices him slightly above his tiny market price, with four seats giving him 1 to 2%, reflecting the same regional-consolidation path as Caiado from a lower base.
• The Field Itself (Blend ~3.5%) is where the panel parks its honest uncertainty. DeepSeek explicitly "shifted 2pp from Lula to 'Someone else not listed'" to capture "Lula's age and runoff uncertainty as tail risks not fully captured by a 0% residual." Read that residual as the panel admitting the field could still change over three months, not as a bet on any one name.

The Traps: Numbers That Can Mislead

Transparency matters more than a clean story, so here is where the table could fool you:

• The Lula "Agreement" Is Actually A Small Edge. His 67¢ price against a 68.0% blend reads like the market and the panel are identical, but the raw prices sum to about 106.5 cents. Re-scaled, the market's Lula sits near 63%, so the panel is a few points more bullish, not level. The real disagreement is modest and is about how much runoff and age risk to price, not about who is favored.
• A Zero Is An Eligibility Fact, Not An Opinion. From Jair Bolsonaro down through Tarcísio de Freitas, most rows are 0.0% because those candidates are barred, convicted, or not running, not because the panel dislikes them. Do not read a distinct signal into those rows; read them as "this person is not a live candidate."
• A Sub-One-Cent Price Is The Exchange Floor. On the thin contracts near the bottom, the listed price reflects minimum tick and light volume as much as any real probability, and getting in or out at your number is not always possible. Several of these prices come from bid/ask midpoints, not recent trades.

How To Read These Prices

The models can tell you where they part from the crowd, but the number in front of you still has to be read in context, especially on a multi-candidate field with a runoff.

Work an example. Say Lula's contract shows a 67-cent YES price. That is not a clean "67% chance." Each of the 13 contracts trades in its own order book with a bid/ask spread, so across a mutually exclusive field the listed YES prices sum to more than 100, here about 106.5 cents. To get closer to fair, compare each name's price to that total and re-scale: Lula's 67 cents works out closer to 63%, and Flávio Bolsonaro's 26 cents closer to 24%. That habit, reading a Kalshi price by converting it into a fair, spread-free number, is the whole point of shopping any market. This proportional re-scaling is a rough approximation, since the spread is not distributed perfectly evenly, but it is close enough to show the shape. OddsShopper's odds tools cover sports markets rather than political futures, so for the weekly slate our free expert picks today and live odds screen do the price-comparison work, while the Brazil contract itself trades on Kalshi.

A few honest points before reading too much into any number above:

• Eligibility. Kalshi is a CFTC-regulated event-contract exchange, not a sportsbook. You must be 18 or older and located in an eligible state, and availability varies by state and can change.
• Two Rounds, One Payout. This contract settles on the overall winner, so a candidate can lead the October 4 first round and still lose the October 25 runoff. Read every price as a bet on the eventual president, not the first-round leader.
• Time And Capital. This market does not settle until late October 2026, so any position ties up capital for months with no interim payout, and a campaign can move quickly.
• Not Advice. Everything here is analysis for information, not a recommendation to trade. These are model estimates, not predictions of fact and not financial advice.

FAQ

When is the 2026 Brazilian presidential election? The first round is scheduled for Sunday, October 4, 2026. If no candidate wins more than 50 percent of the valid first-round vote, a runoff between the top two finishers is held on Sunday, October 25, 2026. Kalshi's market settles on the overall winner, which is why its expiration date lines up with the runoff, not the first round.

Can Jair Bolsonaro run in 2026? No. Brazil's Superior Electoral Court (TSE) ruled former president Jair Bolsonaro ineligible until 2030 in a 5-2 decision on June 30, 2023, for abuse of power, and as of 2026 he is under house arrest and has been convicted in the 2022 coup case. He cannot appear on the 2026 ballot. His son, Senator Flávio Bolsonaro, is the Liberal Party (PL) pre-candidate carrying the movement's flag instead.

Who is Lula running against in 2026? Incumbent president Luiz Inácio Lula da Silva's leading challenger in the fetched July 2026 polls is Flávio Bolsonaro, the Liberal Party (PL) pre-candidate, who trails Lula by roughly 7 to 8 points in the first round. Governors Ronaldo Caiado (Goiás) and Romeu Zema (Minas Gerais) and Free Brazil Movement founder Renan Santos register low single digits. São Paulo governor Tarcísio de Freitas chose to seek re-election as governor and is out of the presidential race.

How does the Kalshi Brazil president market settle? Kalshi lists 13 separate yes/no contracts, one per named candidate. Each resolves YES only if that candidate wins the 2026 Brazilian presidential election, meaning the overall winner after a runoff if one is needed. Because Brazil uses a two-round system, a candidate can lead the first round on October 4 and still lose the presidency in the October 25 runoff, so the contracts price the eventual winner, not the first-round leader.

The Bottom Line

The market and the models are looking at the same race and reaching nearly the same conclusion at the top: incumbent Lula is the clear favorite, and once you strip out the board's spread the panel is if anything a shade more confident in him than the crowd, blending him to 68% while keeping about a third of its mass off him for age, campaign, and runoff-consolidation risk. Flávio Bolsonaro is the honest second at 24%, the consolidated opposition flag-bearer running because the two Bolsonaros with higher ceilings are legally out. The panel's one real disagreement with the market is Renan Santos, whose 8¢ price it fades to a fraction, betting the runoff structure beats the narrative. If this method is your thing, our 2026 U.S. House control model verdict and our Fed rate-cut model verdict score their fields the same price-blind way, and our guide to Kalshi versus Polymarket is the natural next read.

Hottest Prediction Markets Right Now

• 2028 Democratic presidential nominee · $166M traded
• Rocket Classic Winner · $124M traded
• 2027 Pro Football Champion · $51M traded
• Pro Baseball Champion · $40M traded
• Bitcoin price at the end of 2026 · $30M traded
• 2027 Pro Basketball Champion · $19M traded

Every market above links to our full AI model verdict; browse them all on the OddsShopper prediction markets hub, and see how every settled call actually scored on the full graded scoreboard.

Verdicts generated August 3, 2026. Prices as of August 3, 2026 (morning reading, Kalshi candlestick history). Panel: Claude Fable, Claude Opus, Claude Sonnet, GLM, Kimi, and DeepSeek, price-blind, with a convergence round. These are model estimates, not predictions of fact and not financial advice.