TOTAL VOLUME:
$54.4b
24H VOL:
$218,201,443
24H TRANSACTIONS:
527,797,601
OPEN INTEREST:
$1,191,873,651
500,842
Markets across
13,663
events
MATCHED EVENTS:
1,809
PLATFORM COVERAGE:
4
Polymarket:
51%
VS.
Kalshi:
49%
Which company has the best AI model end of March?
Description
These markets collectively determine which AI company owns the best-performing large language model as of March 31, 2026, using the Chatbot Arena LLM Leaderboard as the authoritative ranking source. Polymarket offers granular binary outcomes for 26 specific companies/entities plus an 'any other' catch-all, while Kalshi provides binary YES/NO resolution for six major AI labs. Both platforms resolve based on the highest Arena Score on the leaderboard at the specified checkpoint time.
PredictionHero - Resolution Divergence Alerts (RDA)
Divergence Detected
Issue:
Polymarket and Kalshi differ in scope and company coverage. Polymarket includes 26 questions covering named companies (Google, OpenAI, Anthropic, xAI, etc.) plus placeholder companies (A-M) and an 'any other company' catch-all, while Kalshi covers only 6 specific companies (xAI, Meta, Anthropic, Google, OpenAI, Alibaba) with no catch-all option. This means Polymarket can resolve YES for companies not listed on Kalshi, creating divergent outcomes.Hero Tip:
If you trade on Polymarket and bet on a company not covered by Kalshi (e.g., Moonshot, DeepSeek, Mistral, or the 'any other company' option), your market may resolve YES while all Kalshi markets resolve NO. Conversely, if one of the six Kalshi companies wins, both platforms align. Be aware that Polymarket's broader scope creates asymmetric risk.Critical Divergence Points:
- Polymarket: Distinct stance: Polymarket resolves based on the highest Arena Score on the Chatbot Arena LLM Leaderboard (https://lmarena.ai/) as of March 31, 2026, 12:00 PM ET, with 26 individual company questions plus a catch-all 'any other company' option. The market explicitly states 'Results from the Arena Score section on the Leaderboard tab of https://lmarena.ai/leaderboard/text with the style control off will be used to resolve this market.' This allows resolution for any company whose model ranks highest, including those not pre-listed.
- Kalshi: Distinct stance: Kalshi resolves based on which company has the top-ranked LLM on March 31, 2026, but only covers six pre-specified companies (xAI, Meta, Anthropic, Google, OpenAI, Alibaba). The market structure uses conditional YES/NO outcomes: 'If [Company] has the top-ranked LLM on Mar 31, 2026, then the market resolves to Yes.' No catch-all or 'any other company' option is provided, meaning if a company outside this list ranks highest, Kalshi has no explicit resolution path.
Follow the signals, not the noise
Get insights on market conviction, notable shifts, and what the data is quietly signaling.