Join the ranks of the OddsShopper Hall of Fame
Updated July 23, 2026 · 18 min read by Jake Hari

Jake Hari leads content and growth at OddsShopper and Stokastic, turning the team’s betting data and expert analysis into strategy guides bettors can actually use.

Kalshi prices which lab has the best model on December 31. We asked eight AI models, including the ones being judged, and told them honesty beats loyalty.

BTC trades near $66,000 and Kalshi runs a ladder on when it next touches six figures. Eight AI models priced the whole ladder price-blind, and most say not this year.

The FOMC meets July 28-29 and Kalshi has a 26-cent hike priced in. Eight AI models scored the decision and the full-year cut count without seeing a single price.
Stranger Things spent nearly a decade as Netflix's flagship, and its fifth and final season aired its last episode on New Year's Eve 2025. So why is anyone still paying real money on Kalshi for the chance a new Stranger Things episode shows up before January 1, 2027? The answer lives in the gap between what "the show is over" means to fans and what it might mean to a settlement desk. And as it turns out, that gap has already been put to a live test this year.
| Venue | Kalshi — a CFTC-regulated exchange (18+; availability varies by state, as of July 2026) |
| Ticker | KXMEDIARELEASEST-27JAN01 |
| The Contract | Resolves per the market's official settlement rules: was a new episode of Stranger Things released worldwide before January 1, 2027? |
| Closes | On resolution; scheduled settlement January 1, 2027 |
| Current YES Price | 6¢ ≈ 6% implied (snapshot as of July 20, 2026 — confirm the live price on Kalshi before trading) |
Verify it yourself: Netflix's own Stranger Things 5 release schedule confirms the flagship's three-part 2025 finale (Volume 1 on Nov 26, Volume 2 on Dec 25, the finale on Dec 31), and the animated spinoff Stranger Things: Tales from '85 has its own separate Netflix release history, the exact backdrop this market is pricing.
The bull case starts with the word "episode." Netflix has been loud about keeping this universe alive after the flagship ended, with an animated spinoff already on the way and more projects in development. If any of those releases a first episode worldwide before the deadline, and the rules read "Stranger Things" as the franchise rather than one specific show, Yes wins without a single new minute of the original cast. Even inside the flagship, streamers have made a habit of finale-adjacent extras: specials, epilogues, re-releases with new material, anniversary events. Netflix has an obvious commercial reason to keep the brand warm through 2026 while the spinoff era ramps up, and it controls its own release calendar completely, so a surprise drop needs no runway. The bull is not betting on a resurrection; it is betting on a broad definition and a motivated studio. Hold onto that spinoff argument, because the market has already stress-tested it, and we will get to what happened.
The bear case is the calendar and the closed book. The flagship's fifth season was billed as the end years in advance, the story wrapped, and no mainline episode sits on any announced schedule. Television production runs on multi-year timelines, so a live-action episode nobody has announced does not materialize inside a few months. That routes the entire Yes case through spinoffs and specials, and that is where the fine print bites: a market about "a new episode of Stranger Things" most naturally means the show called Stranger Things, not every project wearing the logo, and an animated series with its own title may simply not count, per the market's settlement rules. Netflix's incentives cut the wrong way too. Spinoffs launch when marketing is ready, not rushed out to beat a New Year deadline. No announced episode, no time to make one, and a narrow reading of the rules.
This market settles on January 1, 2027, and asks one thing: was a new episode released worldwide before that date. "Worldwide" matters, because a territory-limited or theatrical-only release is exactly the kind of edge case that comes down to the fine print. As of mid-July 2026, Yes trades around 6 cents, an implied chance in the mid-single digits. That price is the crowd saying the show is done and nothing qualifying is on the announced calendar, while conceding that a surprise special is not literally impossible. It also tells you how traders read the spinoff question: if an announced franchise project clearly counted, this would trade far higher, so the market is being treated as a mainline-episode question. Watch three things: any Netflix announcement of a special or epilogue, a mainline release landing on the 2026 calendar plus clarity that it qualifies, and otherwise the slow walk toward zero as the year runs out. Release-date markets are their own genre on Kalshi: our Dune Part Three verdict prices a dated studio promise, while our Young Thug and Trippie Redd feature verdict prices the no-calendar version of the same waiting game.
The model panel ran on July 20, 2026, and most of it was working blind on the two releases that reframe this whole question. Here is the verified timeline. Stranger Things' fifth and final season released in three parts in late 2025 — Volume 1 on November 26, Volume 2 on December 25, and the finale on December 31, 2025 — so the flagship's own episodes had already aired before the deadline this market is pricing. Then, on April 23, 2026, Netflix released Stranger Things: Tales from '85, the announced animated spinoff, worldwide. That is a brand-new episode of a show carrying the Stranger Things name, out well before the January 1, 2027 deadline.
And yet the market sits at 6 cents, which tells you more about the contract than any model could. Sit with the puzzle: the mainline Season 5 episodes released before the deadline, back in late 2025, and the spinoff released in April 2026, and still the market never jumped toward YES. If the contract simply paid on any Stranger Things episode arriving before January 1, 2027, it would already read 100 cents. It does not. So the settlement almost certainly counts only a new episode released after the market's own reference point, and it appears to exclude the separately titled animated spinoff on top of that. The public one-line contract does not spell the fine print out, so this is an inference, not a decoding — but it is a strong one, because both obvious YES paths have already fired as live experiments and neither one settled this market. Read the official rules before assuming what still qualifies, and treat the callback to the bull case's spinoff path as closed: it was tested, and the market shrugged.
That reframes the panel. GLM 5.2's 30%, the high outlier that drags the blend up, rests on the assumption that Season 5 was "still in progress with a wide window that plausibly extends into Q4 2026." That premise is simply false: Season 5 finished airing on New Year's Eve 2025. Kimi K3, at 10%, had the sharpest facts (it named the volume air dates and flagged the spinoff) but hedged the spinoff's 2026 release as "unconfirmed" — it is now confirmed, released, and, crucially, ignored by the market. Strip out the outlier built on a wrong premise and the honest read is that the market's 6¢ is well-supported, with the residual chance living entirely in a surprise mainline special that no one has announced.
Market: 6¢ · AI blend: 10%. The blend sits 4 points above the market, but read that gap as a caution flag rather than a signal. As the section above shows, the premium is carried by one outlier (GLM 5.2 at 30%) whose reasoning assumes a Season 5 that had, in fact, already finished airing months earlier.
| Model | YES | Why |
|---|---|---|
| ChatGPT | 4% | The flagship series has concluded, making an unannounced worldwide bonus episode before year-end highly unlikely. |
| Gemini Flash | 2% | Season 5 filming has not yet begun, making a 2026 release highly improbable given typical production timelines. |
| Gemini Pro | 6% | The extremely low market price of 6% strongly suggests an official announcement has already slated the final season's release for 2027 or later. |
| Grok 4.5 | 8% | By mid-2026 with no Season 5 release or firm date announced after 2025 target miss, a 2026 drop is improbable. |
| DeepSeek V4 | 7% | No release by July 2026 implies a 2027 target is likely, but a small chance of a surprise late-2026 drop remains. |
| Kimi K3 | 10% | Season 5 already aired in late 2025, so the 6¢ price implies only post-finale content counts; YES hinges on the announced 2026 animated spinoff releasing and qualifying. |
| GLM 5.2 | 30% | As of July 2026, if Season 5 hasn't released yet, the window before Jan 2027 is narrow and Netflix's silence likely signals a 2027 target. |
| Blended Verdict | 10% | equal-weight mean of 7 models (method always disclosed) |
The models span a 28-point range on this one, and the most telling row is the 30% outlier: it is not a bolder read of the same facts, it is a read of the wrong facts, a Season 5 that its author believed was still unaired.
Model estimates generated 2026-07-20 22:58 UTC. These are model estimates, not predictions of fact and not financial or trading advice. Models are frequently wrong; the market price reflects real traders' money.
The camps disagree less about reasoning quality than about which underlying facts each model is working from: Kimi K3 is the only one to cite the specific Stranger Things Season 5 volume-release schedule (Nov 26/Dec 25/Dec 31, 2025) and the announced 'Tales from '85' animated spinoff (targeted 2026), while GLM 5.2 appears to assume Season 5 is still mid-production with filming starting 'early 2025,' a very different premise from the others. Several models (Gemini Pro, GLM) also lean on the 6-cent price itself as evidence for their conclusion rather than as something to be explained, producing a circularity that pulls estimates toward the price rather than away from it.
The AI read: Kimi K3's case rests on the most specific, checkable facts - a named animated spinoff ('Tales from '85') and explicit Season 5 volume air dates (Nov 26/Dec 25/Dec 31, 2025) - that none of the other six models mention, giving it the strongest evidentiary footing even though its 10% sits mid-pack.
What settles it: An official Netflix announcement - either a confirmed mainline premiere date or explicit confirmation that no further Stranger Things episodes are planned before 2027 - would resolve the disagreement.
ChatGPT (4%): The strongest driver is that Stranger Things’ fifth season was publicly framed as the main series’ final season, with its episodes expected to have been released before July 2026; consequently, another new episode by December 31 would require an unexpected bonus, special, or reversal. For completed prestige streaming series, the six-month base rate of surprise additional episodes is very low, probably below 5%, especially when worldwide availability is required. Announced spinoffs, documentaries, aftershows, or animated projects would not ordinarily count as episodes of Stranger Things itself. A literal reading of the supplied title would resolve YES because episodes released earlier than today were still released before January 1, 2027, but the six-cent price strongly indicates the market intends a new release after its relevant starting point, so I estimate that evident intent. My largest uncertainty is whether Netflix has announced a late-2026 special or whether settlement language counts ancillary Stranger Things programming. The single news item that would most change my estimate is an official Netflix announcement giving a 2026 worldwide release date for a bonus or special episode.
Gemini Flash (2%): The strongest driver for a 'No' resolution is the current production status of Stranger Things Season 5. As of July 2026, filming has not commenced. Given the extensive pre-production, principal photography (which typically lasts 6-9 months for a show of this scale), and lengthy post-production (visual effects, editing, sound mixing), releasing a new episode worldwide before January 1, 2027, is extremely unlikely. Base rates for major Netflix series indicate a minimum of 18-24 months from filming start to release. Even if filming began tomorrow, a 6-month turnaround for a show of this complexity is unprecedented. I am most uncertain about any potential unannounced accelerated production schedule or a surprise 'special' episode that isn't part of Season 5 but still counts as a 'new episode.' The single piece of news that would most change my estimate would be an official announcement from Netflix or the Duffer Brothers stating that principal photography for Season 5 has already been completed.
Gemini Pro (6%): Given the simulated current date of July 20, 2026, the market price of 6 cents is the strongest indicator of the release timeline. Stranger Things Season 5 experienced significant initial delays due to the 2023 Hollywood strikes, pushing the start of filming to early 2024. With the show's extensive post-production requirements-often taking up to a year for heavy VFX-a late 2025 or 2026 release was initially plausible. However, the 6% probability implies that recent developments, likely an official Netflix announcement or severe production delays, have definitively pushed the premiere to 2027 or beyond. Base rates for major VFX-heavy series show that once a show is delayed past its expected window, it rarely moves back up. My greatest uncertainty lies in the exact reason for the delay-whether it is a post-production backlog, extensive reshoots, or strategic scheduling by Netflix. The one piece of news that would most change this estimate is a surprise announcement from Netflix moving the premiere date up to late 2026, perhaps by splitting the season into two parts.
Grok 4.5 (8%): Key driver is the missed 2025 target for Stranger Things Season 5 (final season), with filming having extended and no worldwide episode drop by July 2026, per typical Netflix patterns for tentpole releases that require months of post-production and marketing. Secondary factors include historical delays for the show (e.g., Season 4 gaps) and base rates for prestige TV finales slipping calendar years amid strikes or VFX needs, plus Netflix's preference for Q4 or summer windows now largely passed. Uncertainties stem from my knowledge limits on exact 2025-2026 production updates or secret drop plans. The single most estimate-shifting news would be an official Netflix trailer or premiere date confirmation for fall 2026.
DeepSeek V4 (7%): Key drivers: (1) Production timeline – filming began early 2024 after strikes, likely wrapped by early 2025; massive post-production typically needs 8–12 months, so a 2025 release was plausible, but absence by mid-2026 signals delays. (2) Marketing conventions – Netflix flagships require 3+ months of promotion; if no date is announced by now, a pre-2027 release is improbable. (3) Residual scenarios – a split-volume strategy or surprise Christmas drop could still enable a late-2026 release. Base rate for final seasons delayed 4+ years is low. I'm most uncertain about whether post-production setbacks or strategic scheduling have definitively pushed it to 2027. One piece of news: an official 2027 release date announcement would drop my estimate near zero; a surprise trailer would spike it.
Kimi K3 (10%): Season 5's announced schedule (Volume 1 Nov 26, Volume 2 Dec 25, finale Dec 31, 2025) means a literal reading already resolves YES, so the 6¢ price is coherent only if the market counts episodes released after those air dates or after market creation; per instructions I estimate under that evident intent and flag the discrepancy. The dominant YES path is the animated spinoff 'Stranger Things: Tales from '85,' announced April 2025 with a reported 2026 target: Netflix typically confirms dates two-to-four months ahead, and I know of no confirmed date, leaving a Q4 2026 premiere plausible but far from assured; animation schedules slip often (announced-year streaming releases hold roughly 60-70%). Resolution-source risk cuts hard: rules keyed to the main series' episode list would exclude a spinoff entirely. There is no Season 6, the live-action spinoff is years away, and specials likely don't qualify. I'm most uncertain about post-announcement news (e.g., a delay to 2027) that the low price may already reflect. Most decisive news: Netflix confirming a 2026 premiere date for Tales from '85.
GLM 5.2 (30%): Stranger Things Season 5 (the final season) began filming in early 2025. The show is VFX-heavy with historically long post-production cycles; Season 4 took roughly two years from filming start to release. If filming wrapped around mid-2025, 12-18 months of post-production would place release roughly mid-2026 to early 2027. The market price at 6 cents strongly suggests that by July 2026, either no release has occurred and no firm late-2026 date has been announced, or delays have been confirmed. Netflix typically promotes major releases months in advance, so absence of a confirmed 2026 date by July is a negative signal. However, Netflix could still drop episodes in late 2026 (Oct-Dec) with a shorter promotional window, or release in two volumes as with Season 4. My key uncertainty is whether filming wrapped on schedule and whether Netflix has made any official release-date announcements I'm unaware of. The single piece of news that would most change my estimate: an official Netflix announcement of a 2026 release date, which would push my estimate above 0.70.
A prediction-market price is not a vibe; it is a probability with a dollar sign in front of it. YES at 6¢ is the market's shorthand for "about a 6% chance," the same implied-probability math you would run to de-vig a two-way sportsbook line. What sharp bettors like about a venue such as Kalshi is that the price sits close to no-vig. You are not paying the 4-5% hold a book bakes into a two-way market, so 6¢ is a cleaner read on true probability than a lopsided moneyline would be. If you have never turned a market price into a percentage before, our primer on how prediction markets work walks through it step by step.
The first move is just translating the price. Six cents means about a 6% chance, and if turning market quotes and American-style prices into clean probabilities is not yet second nature, OddsShopper's betting calculators do that conversion for you. The next move is the one you would make on any number: compare that implied probability to your own estimate, and only act when the two diverge. The on-topic tool for that step is OddsShopper's EV calculator, which turns any price-plus-estimate into an expected-value figure in one shot — feed it the 6¢ price and a probability you actually believe, and it does the arithmetic below for you. The catch, as the next section shows, is that the model blend does not give you a probability you should actually believe. And before you trade any political, culture, or event market, confirm it is available where you live; we keep a running guide to whether prediction markets are legal in your state.
The one number I keep coming back to is not the 10% blend but the 6¢ price, and the puzzle underneath it. Two different Stranger Things releases have already landed before this market's deadline, the mainline finale in December 2025 and the animated spinoff in April 2026, and neither flipped it toward YES. That tells me the contract is quietly keyed to something narrower than "any new episode," and it is the whole reason a real-money market can sit at six cents on a show everyone assumes is finished. The bet was never the models. It is whether Netflix drops a qualifying release in the roughly five months left on the clock.
Say you take the 10% model blend at face value and set fair value there. Here is the napkin math a bettor runs before risking a dollar:
| Input | Value |
|---|---|
| YES Price | 6¢ (risk 6¢ to win 94¢) |
| The Model Blend As Fair Value | 10% |
| Break-Even Probability | 6% |
| Expected Value Per Contract | (0.10 × $0.94) − (0.90 × $0.06) = +$0.04 |
On those inputs, every YES contract shows about four cents of theoretical edge: positive, but thin, and entirely dependent on trusting a model blend over the market. Now flip the estimate the honest way. Drop GLM 5.2's 30% — the number built on a Season 5 that had already finished airing — and the remaining six models average roughly 6%, right on top of the market price. On that reading the expected value is essentially zero, and any Kalshi fees drag it negative from there. That is the real read: the apparent edge is an artifact of one outlier working from a wrong fact, not a live opportunity. But running the number instead of trading the headline is the whole discipline, and it is exactly what the 6¢ price is inviting you to do.
Is a spinoff or special enough to settle this market Yes?
The animated spinoff Tales from '85 released worldwide on April 23, 2026, and the market did not resolve Yes — which strongly implies the settlement rules key on the mainline live-action series, not every project wearing the logo. Read the rules before assuming any spinoff or one-off special qualifies.
Is there a hard deadline for a release to count?
Yes. The episode must be released worldwide before January 1, 2027, and the market settles that day. The worldwide requirement means a limited or single-region release is a rules question, not an automatic Yes.
Is Kalshi legal in my state?
Availability varies by state and changes over time, so check your eligibility directly on the platform before trading. You must be 18 or older. Accurate as of July 2026.
Are the model verdicts advice?
No. They are model estimates of how a market might resolve, not financial advice. Any decision to trade or abstain is yours, based on your own judgment and risk tolerance.