Kalshi prices as of Thursday, September 10, 2026, 9:28 AM EDT, the board every model read; the Bet now links open the live market, and the ask may have moved.
As of Thursday, September 10, 2026, 7 AI models have each picked three Kalshi player-prop contracts for NFL Week 1 from the same written playbook and the same data. On Kalshi a prop is a yes-or-no contract at a fixed line, priced in cents: 40 cents means a 40% chance. This board is separate from the sportsbook panel; a model was free to take anything Kalshi lists, and the games already played or kicking off before every model could run (NE at SEA, SF at LA) were left off. Nobody told them who to pick, no model saw another's picks, and every contract is graded against the box score, with each model's profit tracked in units all season.
Stay ahead of the markets.
Daily insights and expert picks on Kalshi, Polymarket, and what's moving markets.
Free forever. Unsubscribe anytime.
The Read
Given a board of yes-or-no contracts instead of lines, the models went where the sportsbooks rarely post: receptions, completions, and the no side of a number the market seems sure about. Two of them landed on the same rookie tight end's catches without seeing each other's work. The contracts are below in each model's own words, with a link to the market. The record starts Monday.
How it works
Every model gets two things: Dave Loughran's playbook for researching props, which says what to look at, what counts and what to ignore, and a data packet covering the whole slate from the Kalshi board. Each model picks three, writes its reasoning for each, and hands back the list. How we keep score: every bet is to win one unit. A yes bought at 40 cents pays 60 on a win, so it risks 1 unit to win 1.5; a yes at 60 cents risks 1.5 units to win 1. Wins and losses add up week to week.
The Standings
| Model | Record | Units | Weeks |
|---|---|---|---|
| Claude Fable 5.1 | 0-0 | +0.00 | 1 |
| OpenAI GPT (codex) | 0-0 | +0.00 | 1 |
| Google Gemini Pro | 0-0 | +0.00 | 1 |
| xAI Grok | 0-0 | +0.00 | 1 |
| Claude Opus 5 | 0-0 | +0.00 | 1 |
| Moonshot Kimi | 0-0 | +0.00 | 1 |
| Claude Sonnet 5 | 0-0 | +0.00 | 1 |
Nothing is graded yet; the first grades post after this week's games.
The Board
Kalshi in cents; the ask is what a buyer pays.
| Model | Contract | Side | Price |
|---|---|---|---|
| Claude Fable 5.1 | Alec Pierce 50+ receiving yards | Yes | 47¢ |
| Claude Fable 5.1 | Harold Fannin Jr. 4+ receptions | Yes | 59¢ |
| Claude Fable 5.1 | Lamar Jackson 18+ completions | No | 51¢ |
| OpenAI GPT (codex) | Brock Bowers 6+ receptions | Yes | 55¢ |
| OpenAI GPT (codex) | Alec Pierce 50+ receiving yards | Yes | 47¢ |
| OpenAI GPT (codex) | Harold Fannin Jr. 4+ receptions | Yes | 59¢ |
| Google Gemini Pro | Dalton Kincaid 4+ receptions | Yes | 47¢ |
| Google Gemini Pro | Jonathan Taylor 3+ receptions | Yes | 58¢ |
| Google Gemini Pro | Quentin Johnston 4+ receptions | Yes | 44¢ |
| xAI Grok | Javonte Williams 17+ rush attempts | Yes | 58¢ |
| xAI Grok | Cade Otton 4+ receptions | Yes | 41¢ |
| xAI Grok | Quinshon Judkins 60+ rushing yards | No | 57¢ |
| Claude Opus 5 | Cade Otton 4+ receptions | Yes | 41¢ |
| Claude Opus 5 | Marvin Harrison Jr. 4+ receptions | No | 44¢ |
| Claude Opus 5 | Harold Fannin Jr. 4+ receptions | Yes | 59¢ |
| Moonshot Kimi | Bucky Irving 60+ rushing yards | Yes | 50¢ |
| Moonshot Kimi | Rashee Rice 6+ receptions | Yes | 49¢ |
| Moonshot Kimi | Charlie Kolar 2+ receptions | No | 43¢ |
| Claude Sonnet 5 | CeeDee Lamb 80+ receiving yards | Yes | 48¢ |
| Claude Sonnet 5 | Stefon Diggs 6+ receptions | Yes | 23¢ |
| Claude Sonnet 5 | De'Von Achane 5+ receptions | Yes | 43¢ |
Claude Fable 5.1
1. Alec Pierce 50+ receiving yards, yes at 47 cents on Kalshi. Bet now on Kalshi
Baltimore's pass defense leaked in exactly the way Alec Pierce attacks. The Ravens allowed 263.4 passing yards per game (second-most in the league), were thrown on at the fifth-highest rate, and gave up a 45.8% completion rate on deep balls — fourth-worst — while facing 4.9 deep attempts per game, the most in football. Pierce is the league's most extreme deep receiver: a 19.0 average depth of target, 25 deep targets, a 36.4% share of Indianapolis's air yards, and 11.94 yards per target.
2. Harold Fannin Jr. 4+ receptions, yes at 59 cents on Kalshi. Bet now on Kalshi
No contract this week has more separate signals landing on the same square. Fannin caught 4 or more passes in 11 of 16 games as a rookie on 108 targets and a 20.6% target share — and that was while splitting the room with David Njoku, whose 48 targets left for the Giants with no veteran replacement added. The scheme fit is the sharpest part: Cleveland threw to its tight ends on 36.8% of attempts against zone coverage, the highest share in the league, and Jacksonville played zone 78.6% of the time, second-most.
3. Lamar Jackson 18+ completions, no at 51 cents on Kalshi. Bet now on Kalshi
This is a bet on Baltimore staying Baltimore. The Ravens passed at the lowest neutral-script rate in the league (47.3%) with the second-lowest pass rate over expected, averaged just 31.4 dropbacks per game, and Lamar Jackson reached 18 completions in only 4 of 13 starts — under it 69% of the time. Nothing this offseason argues for more volume: the new play-caller, Declan Doyle, comes from Ben Johnson's Chicago staff, which ran the ball at the third-highest rate in 2025; Derrick Henry still owns a 68% carry share at 18 carries a game; and the departed targets (DeAndre Hopkins and Isaiah Likely, 21% of the team's) were replaced by a rookie, Ja'Kobi Lane.
OpenAI GPT (codex)
1. Brock Bowers 6+ receptions, yes at 55 cents on Kalshi. Not graded (zero units): Brock Bowers was ruled out Wednesday, September 9 (meniscus trim); the data the models were given was built before that report, so the pick is voided at zero units.
Bowers is the best blend of role, market type and matchup. He drew 7.2 targets per game last season, had a 23.6% Sumer target share, ran hot near the goal line and played 96.5% of the snaps over his final four games. Las Vegas also has Kirk Cousins now, which points toward timing throws, and Kubiak’s last two offenses ranked top-8 in red-zone TE share.
2. Alec Pierce 50+ receiving yards, yes at 47 cents on Kalshi. Bet now on Kalshi
Pierce has the cleanest price-to-line edge of the group: a 50+ receiving-yards contract at 47 cents, with the sportsbook number sitting at 45.5. He cleared 50 yards in 10 of 15 games last season while playing an 85.9% snap share, and his profile fits a downfield matchup: 19.0 aDOT, 36.4% air-yards share, 25 deep targets and 11.94 yards per target.
3. Harold Fannin Jr. 4+ receptions, yes at 59 cents on Kalshi. Bet now on Kalshi
Fannin is the strongest short-area volume play after Bowers. He had a 0.240 targets per route run rate, 108 targets, 72 catches and cleared four receptions in 11 of 16 games last season. Cleveland also moved into Todd Monken’s offense, and Monken’s Ravens ranked fourth and fifth in TE target share over the last two seasons while ranking first in red-zone TE share both years.
Google Gemini Pro
1. Dalton Kincaid 4+ receptions, yes at 47 cents on Kalshi. Bet now on Kalshi
Opportunity begins with Kincaid commanding a 14.6% target share in this offense last season, yielding 4.2 targets per game despite missing time. He operates as a central chain-mover for Buffalo, and his specific matchup this week is a massive outlier. Houston relies heavily on zone coverage, deploying it on 74.7% of their defensive snaps.
2. Jonathan Taylor 3+ receptions, yes at 58 cents on Kalshi. Bet now on Kalshi
Taylor commanded an 80.5% carry share last season but also ran a route on most passing downs, earning 56 targets. Vacated passing-down snaps consolidate the backfield target tree directly onto Taylor. He faces a Baltimore defense that operates as a clear pass funnel, ranking 12th-worst in pass EPA allowed but 25th in rush EPA allowed, which naturally pushes offenses to the air.
3. Quentin Johnston 4+ receptions, yes at 44 cents on Kalshi. Bet now on Kalshi
Quentin Johnston is poised for a massive volume leap. Looking at the roster, the departure of Keenan Allen vacates a staggering 122 targets—23% of the Chargers' entire passing offense—leaving Johnston to inherit a premier role. He already commanded 6.5 targets per game last year.
xAI Grok
1. Javonte Williams 17+ rush attempts, yes at 58 cents on Kalshi. Bet now on Kalshi
The Giants’ 2025 defense was the softest run unit in the league by a mile: first in rush EPA allowed, first in yards per carry (5.30), first in explosive-run rate allowed, and second in rush yards per game. That is a volume matchup, not a yards bet dressed up as one. Williams handled 60.6% of Dallas’s carries (15.8 per game, 9.2 on neutral script) and 66.7% of the red-zone work, so the opportunity is already baked into his role.
2. Cade Otton 4+ receptions, yes at 41 cents on Kalshi. Bet now on Kalshi
Cincinnati allowed the most tight-end production in the league last year — 9.7 targets, 6.8 receptions and 84.9 yards per game to the position, with Al Golden still coordinating that defense. Otton is the every-down answer in Tampa: 93.5% of offensive snaps, 5.5 targets per game, 72% catch rate, and 4-plus catches in 8 of 15. Against zone (Cincinnati’s 67.7% diet) he posted a 17.4% target rate and 1.4 yards per route run on 333 routes.
3. Quinshon Judkins 60+ rushing yards, no at 57 cents on Kalshi. Bet now on Kalshi
Jacksonville allowed the fewest rushing yards in the league (70.8 per game, #32) at 3.63 yards per carry, and the front that produced those ranks is mostly intact. Cleveland is an 8.5-point road dog in a 40.5 total, so the script that feeds a lead back — long, tied drives — is the one least likely to show up. Judkins still handled 59.8% of Cleveland’s carries (16.5 per game) and cleared 60 yards in 7 of 14, but that distribution came against a random draw of run defenses, not the stingiest unit in football while trailing.
Claude Opus 5
1. Cade Otton 4+ receptions, yes at 41 cents on Kalshi. Bet now on Kalshi
No defense gave more to tight ends than Cincinnati last year. The Bengals allowed 9.7 targets, 6.8 catches and 84.9 yards a game to the position, gave up 16 touchdowns to it, and sent 31% of all the targets they faced to tight ends. They led the league in every one of those.
2. Marvin Harrison Jr. 4+ receptions, no at 44 cents on Kalshi. Bet now on Kalshi
Kalshi has the yes at 59 cents with a 56-cent bid, which gives Harrison about a 57% chance of catching four passes. The no costs 44 cents, or a 44% chance. His 2025 season doesn't back that price.
3. Harold Fannin Jr. 4+ receptions, yes at 59 cents on Kalshi. Bet now on Kalshi
Cleveland's tight end targets now go to Harold Fannin Jr. alone. David Njoku left for the Chargers and took 48 targets with him. Cedric Tillman and Jerome Ford are also gone, and the three of them account for 23% of last year's targets.
Moonshot Kimi
1. Bucky Irving 60+ rushing yards, yes at 50 cents on Kalshi. Bet now on Kalshi
2. Rashee Rice 6+ receptions, yes at 49 cents on Kalshi. Bet now on Kalshi
3. Charlie Kolar 2+ receptions, no at 43 cents on Kalshi. Bet now on Kalshi
Claude Sonnet 5
1. CeeDee Lamb 80+ receiving yards, yes at 48 cents on Kalshi. Bet now on Kalshi
2. Stefon Diggs 6+ receptions, yes at 23 cents on Kalshi. Bet now on Kalshi
3. De'Von Achane 5+ receptions, yes at 43 cents on Kalshi. Bet now on Kalshi
Bottom Line
7 models, 21 contracts, one board. The standings start at zero and fill in Monday. The picks are the models' own, priced at the ask they saw; nothing here is a Dave Loughran pick unless it also appears on one of his pages.
Follow Dave on X: @Loughy.
Hero illustration: OddsShopper, in the house collage style. CeeDee Lamb photo by All-Pro Reels, licensed CC BY-SA 4.0; Stefon Diggs photo by Keith Allison from Hanover, MD, USA, licensed CC BY-SA 2.0; photos cropped, toned, and composited.
Prices quoted are live Kalshi market prices as of Thursday, September 10, 2026, 9:28 AM EDT, and will keep moving until these markets settle after the Week 1 games. Kalshi is a CFTC-regulated event-contract exchange (18+; availability varies by state). OddsShopper covers prediction markets as an independent analyst: any model or panel figures referenced are model estimates, not predictions of fact and not financial advice. Betting involves risk; contracts can go to zero. If you or someone you know has a gambling problem, call 1-800-GAMBLER.
