When will AI score above 170 on Epoch Capabilities Index?

Kalshi Tech › AI Closes Jul 1, 2027 · $558 traded · $438 at stake · prices read 48 min ago

The odds

Dec 2, 2026even chance by

Before Dec 1, 202649%
Before Jan 1, 202782%▲ 2
Before Feb 1, 202782%
Before Apr 1, 202794%
Before Jul 1, 202796%

How the odds have moved

Before Dec 1, 2026: 46% on Oct 20, 49% now.

Each outcome's chance over time on Kalshi. The dots along the bottom are headlines about its subject: point at one to read it.

Every outcome

Outcome	Chance	Day	Bid	Ask	Pays	Contracts
Before Dec 1, 2026	49%		1¢	99¢	$1.01	200
Before Jan 1, 2027	82%	▲ 2	79¢	84¢	$1.19	541
Before Feb 1, 2027	82%		78¢	86¢	$1.16	24.00
Before Apr 1, 2027	94%		89¢	98¢	$1.02
Before Jul 1, 2027	96%		93¢	99¢	$1.01

"Pays" is what a $1 stake returns if that outcome happens, at the price shown (before any fees). Bid and ask are the best prices to sell and buy "Yes".

The news: Epoch Capabilities Index

Search ›

How could the US and China verifiably slow AI? — LessWronglesswrong.com · 3 days agoAn overview of the game theory of why a unilateral slowdown doesn't work, an overview of various unilateral and bilateral verification methods, and a short scenario starting from after a warning shot combining these tools. AI development…
How to Increase America’s AI Lead Over China | IFPInstitute for Progress · 6 days agoThe US’s top AI models lead top Chinese models by only about 6.5 months. This is despite US AI companies having significantly better compute access and financing. Over the next year, the size of this gap is important for: The size of the…
Anthropic Warns Small AI Models May Soon Self-Replicate on IoT DevicesDEV Community · 7 days agoAnthropic’s... Tagged with anthropic, aisafety, iot, cybersecurity. #anthropic-warns-small-ai-models-may-soon-selfreplicate-on-iot-devices Anthropic Warns Small AI Models May Soon Self-Replicate on IoT Devices #key-experimental-findings…
RedwoodBench: Can Frontier Models Imitate Conceptual AI Safety Researchers? — LessWronglesswrong.com · 9 days agoThis is a two week sprint project done as part of the Redwood Research internship. Hasan was the main IC and proposed the project, Caspar supervised. We find that generally more capable models perform better on our benchmark. RedwoodBench…
Google's Gemini 4 Argon is closer to the AI frontier than any Google model in monthsThe Decoder · 10 days agoGoogle Deepmind is the second major AI lab to pass 95 percent on the ARC-AGI-3 benchmark. But the results depend heavily on the testing method, and Argon trails rivals in math and software engineering, according to Epoch AI. In the…
Epoch InnovationEval: Can AI Rediscover a Real ML Breakthrough?explainx.ai · 15 days agoAI R&D Automation, Epoch AI, Recursive Self-Improvement, AI Evaluation, Research Part of AI Safety and Alignment Epoch AI tested GPT-5.6 Sol and Claude Fable 5 on rediscovering SDPO. Best result: about 35% of its gains. What InnovationEval…
Claude Opus 5.5 Tops the Epoch Capabilities Index With 167 — What the Score Meansexplainx.ai · 19 days agoClaude, Anthropic, AI Benchmarks, Model Comparison, Epoch AI Part of Anthropic and Claude Claude Opus 5.5 scores 167 on Epoch AI's Capabilities Index, 0.84 ahead of GPT-6 Astra. What ECI measures, where Astra still wins, and how to use it…
Fable 5.1 vs Claude Opus 5.5: Which One Do You Actually Needexplainx.ai · 23 days agoComparison, Claude, Fable, Opus 5.5, Benchmarks Opus 5.5 beats Fable 5.1 on every Anthropic benchmark at lower cost. So why does Reddit still say Fable wins? The numbers and the reaction. The reaction split cleanly into two camps, and both…

Related markets

All ai ›

Tech › AI

Which company has the best AI model end of October?

85%Google

Google85%▼ 4
Anthropic13%▲ 5
OpenAI1%▲ 1

18 more

Meta<1%
Moonshot<1%
Alibaba<1%
Baidu<1%
Z.ai<1%
Xiaomi<1%
SpaceXAI<1%
DeepSeek<1%
Amazon<1%
ByteDance<1%
Nvidia<1%
Thinky<1%
MiniMax<1%
Mistral<1%
Meituan<1%
Tencent<1%
StepFun<1%
Microsoft<1%

Tech › AI

Next Fable Model (5.2+) released on...?

69%No release by October 31

No release by October 3169%▲ 5
October 2716%▼ 3
October 286%▲ 2

11 more

October 264%
October 294%▼ 1
October 301%▼ 1
October 201%▼ 12
October 31<1%▼ 1
October 19<1%▼ 2
October 21<1%▼ 5
October 22<1%▼ 14
October 23<1%▼ 4
October 24<1%
October 25<1%▼ 1

Tech › AI

Which company has best AI model end of 2026?

54%Anthropic

Anthropic54%▲ 4
Google33%▼ 6
OpenAI6%

12 more

Meta1%
xAI1%
Alibaba<1%
Moonshot<1%
DeepSeek<1%
Z.ai<1%
Mistral<1%
Microsoft<1%
ByteDance<1%
Meituan<1%
Amazon<1%
Baidu<1%

Tech › AI

Next Claude Haiku Model (4.6+): Text Arena Debut?

1,410median

1,3801,480

Median 1,410Middle half 1,405 – 1,415

Tech › AI

Next Fable Model (5.2+) released by...?

Nov 19, 2026even chance by

October 21<1%▼ 12
October 22<1%▼ 14
October 24<1%▼ 1

2 more

October 3131%▼ 4
December 3194%▼ 1

Tech › AI

Louisiana enacts data center moratorium by...?

34%December 31, 2028

December 31, 20264%
June 30, 202711%
December 31, 202714%

1 more

December 31, 202834%▲ 10

Odds are traders' prices on Kalshi, Polymarket, ForecastEx and Crypto.com, not forecasts by 4Conservative. Prices read from the venues 3 min ago. Prices may be delayed. Money traded on Kalshi and ForecastEx is estimated from contracts. Not betting or investment advice.

About this market

VenueKalshi

Settles onEpoch AI

Trading closesJul 1, 2027 10:00 AM ET

Traded$558 (765 contracts)

This week$103

At stake$438

Followed here sinceOct 19, 2026

The rules

If any AI model company records above 170 on Epoch AI's Epoch Capabilities Index (General ECI) (https://epoch.ai/eci?view=graph&tab=leaderboard) after market issuance and before December 1, 2026 at 12:00 AM ET, then the market resolves to Yes.

The full rules are on the market's page at Kalshi.

Reading the odds

A prediction market trades contracts that pay $1 if something happens. Its price is the market's chance: a contract at 62¢ means traders give it 62%. Each number here is the midpoint between the best bid and offer, or the last trade when they are more than ten cents apart.

For a question with many answers (who wins, which range a number lands in), the chances are scaled to add up to 100%. A price ladder ("above 4.25%, above 4.50% ...") gives a whole distribution: its median is where traders see an even chance either way.

Kalshi is a U.S. exchange regulated by the CFTC; its money is counted in contracts, shown here as dollars at their price. Polymarket counts dollars. ForecastEx is Interactive Brokers' exchange, regulated by the CFTC; its contracts too are shown as dollars at their price. Crypto.com lists its markets on Crypto.com | Derivatives North America, a U.S. exchange regulated by the CFTC; it reports the dollars traded. Where more than one asks the same question, each is shown, never averaged.

Prefer betting odds? Switch to American (+150) or decimal (2.50) odds at the top of the page.

Prediction-market prices are what traders pay for each outcome on Kalshi, Polymarket, ForecastEx and Crypto.com, read as the market's implied chance. They are not forecasts by 4Conservative, can move fast and can be wrong; they are read every few minutes and may be delayed. Money traded on Kalshi and ForecastEx is estimated from contracts at their price. News comes from the 4Conservative index; prices of stocks, rates and coins from 4Conservative Finance. Nothing here is betting, financial or investment advice. Terms of use. How to use Predictions