A9 PREDICT

A9 Predict

Changelog

Changes associated with this build.
v0.2.0Build identifier supplied

Build date · Eastern: Oct 25, 2026, 1:41:30.352 PM EDT

Technical build details
Commit
ef695042b6cc7b8d0132bc9eb2a0a1890d123d57
Version source
Running build metadata

Build metadata does not confirm production deployment or service health.

v0.2.0 This build

Keep this round's issued prediction readable, with a separate fresh entry comparison when matching evidence is available.

Added

  • One-contract entry estimate from matching fresh forecast, quote, depth and timestamped series fees; no orders are placed.
  • Current Kalshi round, threshold, close countdown and timestamped public quotes, separate from immutable A9 forecasts.
  • Eastern-day count of unique rounds with acknowledged forecasts and saved settlement status; real trades remain unavailable without fills.

Fixed

  • Saved history stays historical, including records from an open round; it does not show a current pick or start quote refreshes.
  • An issued prediction stays visible until its round closes, with its original probability and an explicit stale-input label after freshness expires.
  • Invalid reference timestamps, conflicting records and different-round quotes cannot supply a selection or entry comparison.
  • Forecast-only pilots can use the approved budget while learning stays off.
  • Quotes retire when stale, closed, mismatched or incomplete; an old forecast is never relabeled as current.
  • Bounded response parsing and shared in-flight public reads prevent simultaneous viewers multiplying upstream work.

Changed

  • Target prices spell out At least versus Above from the contract comparison; original Kalshi wording and rules remain in Details.
  • Untracked no-pick counts and the absence of a trading connection are labeled explicitly instead of appearing as a temporary data outage.
  • Canonical ACE9 Living Glass theme, compact historical evidence, and Eastern Time throughout.
  • Accessible version dialog groups Added, Fixed and Changed notes for this exact build.
  • Page reads do not start forecast collection, learning or order execution.

Tracked question

Will Xiaomi have the best Chinese AI model at the end of September 2026?

A9 gives this a 82% chance of happening. That's somewhat less likely than the betting market thinks (86%) — a 4-point disagreement. A9 researched this on the web before forecasting.

Market 86%Polymarket
A9 Predict 82%high confidence

Jev on its own said 69%

-4 pts vs market
Leans YES

Why A9 thinks this

The evidence, separated.

Web research · quality 70/100

As of the most recent Arena.ai Text Arena (Overall) snapshot visible on the site (page date Sep 25, 2026), Xiaomi’s mimo-v2.6-pro is the highest-ranked model among the primarily Chinese companies listed on that leaderboard (rank #9, score 1491±9), ahead of Alibaba, Z.ai, Baidu, Moonshot and others. Xiaomi also released the MiMo-V2.6 series in late September 2026 and is running large RL training/validation activity that could increase Arena preference evidence. However, Arena rankings are dynamic and depend on ongoing head-to-head human-preference votes; Xiaomi’s current leading position among Chinese labs rests on relatively modest vote totals for its top entry, making the ranking volatile in the short three-to-four day window before the market’s check time (Sep 30, 2026 12:00 PM ET). Therefore the current evidence supports Xiaomi as the likely top Chinese company on the Text Arena (Overall) leaderboard today, but the outcome can reasonably change before the resolution time due to additional votes, competitor releases/updates, or leaderboard availability.

For

Arena.ai Text Arena (Overall, style control off) snapshot (dated Sep 25, 2026 on the page) lists Xiaomi mimo-v2.6-pro at Rank #9 with score 1491±9 — this is the highest-ranked model on that leaderboard that is clearly identified as from a primarily Chinese company.

Xiaomi publicly released and open-sourced the MiMo-V2.6 series (Pro and Flash) in late September 2026; the new release is being run with large RL training runs and Xiaomi publicized improvements in several benchmarks, which could increase Arena preference votes and Elo over the days after release.

Xiaomi already has multiple entries on Arena’s Text Arena (e.g., mimo-v2.5-pro, mimo-v2.6 variants) indicating active adoption and sufficient exposure on the platform (which helps accumulate votes quickly).

Arena’s leaderboard is driven by head-to-head human-preference votes collected on the platform; recent model releases that get rapid usage/votes can move up the ranking within days, which is consistent with Xiaomi’s recent release and visible vote counts for its models.

Against

Several other primarily Chinese labs already have competitive models on the same leaderboard (Alibaba qwen3.8-max at Rank #17 with 1480±5, Moonshot/Kimi K3 at Rank #23 with 1476±5, Z.ai GLM-5.3 at Rank #25 with 1472±6, Baidu Ernie-5.1 at Rank #31 with 1468±5). Any of these could surpass Xiaomi by Sep 30 if they get more votes or publish updates.

Xiaomi’s top Arena entry (mimo-v2.6-pro) currently (Sep 25 snapshot) has a relatively small vote count (4,026 votes) compared with some other models; smaller sample sizes make ranks more volatile and easier to overtake in a short time window.

Top overall leaderboard positions are dominated by non-Chinese labs (Anthropic, Google, etc.), showing the competitive pressure and making it possible that another Chinese lab releases or tweaks a model before Sep 30 that outperforms Xiaomi on Arena.

Arena rankings can move quickly due to concentrated voting campaigns, rapid adoption by third-party platforms, or changes to which model versions are being voted on — all of which could push a competitor ahead of Xiaomi between Sep 26 and Sep 30.

Unknown

How voting and adoption on arena.ai will evolve between now (Sep 26, 2026) and the exact resolution time (Sep 30, 2026 12:00 PM ET): specifically, whether Xiaomi’s newly released MiMo-V2.6-Pro accumulates enough additional human-preference votes (or whether a competitor gains votes or releases an improved model) to change the ordering of Chinese companies on the Text Arena (Overall) leaderboard.

What would change A9's mind

  • A competing Chinese lab (e.g., Alibaba, Moonshot, Z.ai, Baidu, DeepSeek, ByteDance) releases a new model update or a new model is added to Arena and accumulates enough votes to rank above Xiaomi by Sep 30 12:00 PM ET.
  • Large, concentrated voting campaigns or rapid integration of a competitor model into high-traffic platforms increases its vote count and Arena Elo past Xiaomi’s top Chinese-ranked model.
  • Xiaomi’s mimo-v2.6-pro collects few additional votes (or loses relative preference in head-to-heads) causing its Arena rank/score to drop below another Chinese model before the check time.
  • Arena.ai experiences partial data changes, reclassification, or a scoreboard correction for model identities/sources that reassign a top-ranked model’s lab to a different (or non-Chinese) owner affecting the highest-placed primarily Chinese company.
  • The Arena leaderboard is down at the check time (the market rule says the market remains open until the leaderboard comes back online).

Probability history

A9 never rewrites the past.

MarketA9Me
82%Initial read

As of the most recent Arena.ai Text Arena (Overall) snapshot visible on the site (page date Sep 25, 2026), Xiaomi’s mimo-v2.6-pro is the highest-ranked model among the primarily Chinese companies listed on that leaderboard (rank #9, score 1491±9), ahead of Alibaba, Z.ai, Baidu, Moonshot and others. Xiaomi also released the MiMo-V2.6 series in late September 2026 and is running large RL training/validation activity that could increase Arena preference evidence. However, Arena rankings are dynamic and depend on ongoing head-to-head human-preference votes; Xiaomi’s current leading position among Chinese labs rests on relatively modest vote totals for its top entry, making the ranking volatile in the short three-to-four day window before the market’s check time (Sep 30, 2026 12:00 PM ET). Therefore the current evidence supports Xiaomi as the likely top Chinese company on the Text Arena (Overall) leaderboard today, but the outcome can reasonably change before the resolution time due to additional votes, competitor releases/updates, or leaderboard availability.

a9-predict-0.3.0 · 8f4cc6de15f9606d8f848c766f3a907e9e50f0e33d02533ef2227ad937e64ed9