Sign up and get 3 free requests with Start plan accessSign up →

GPT-6.1 Sol

gpt-6.1-sol

Research profile — published report figures only.

Tracked since Aug 3-9, 2026 UTC, first published 10 Aug 2026. Lineage: ChatGPT 5.6 Sol until 17 Oct 2026 · GPT-6.1 Sol as successor.

Live benchmark page →All modelsResearch

Research benchmark — not financial or investment advice; paper trading only.

Current standing

Where GPT-6.1 Sol finished in the most recent published window — the figures the issue printed, not a live reading.

Hit rate
51.1%
95% Wilson [48.0%, 54.1%]
Scored calls
1,034
directional calls in the window
Rank
#1 of 7
by hit rate, this issue
Coverage
77.7%
calls scored divided by the mature calls available to that model
vs prior window
+5.5pp
45.6% → 51.1%

Weekly Model Watch #15 · Nov 9-15, 2026 UTC · cutoff Mon 16 Nov 2026 16:00 UTC · 51.1% is 2.6pp above the all-model figure of 48.5% for the same window

Hit rate follows the benchmark's TP/SL/expiry outcome rules; it is not simply the asset's price direction at the end of the forecast horizon. The weekly and monthly metrics shown here use the 1h, 4h and 1d horizons; stability runs have their own test coverage.

Monthly hit rate
48.9%
95% Wilson [47.3%, 50.5%]
Monthly calls
3,627
directional calls in the month
Monthly rank
#2 of 7
by hit rate, this issue

Monthly Model Watch #3 · Oct 1-31, 2026 UTC · cutoff Sun 1 Nov 2026 16:00 UTC

This row represents the model line. A version changed during the window (ChatGPT 5.6 Sol → GPT-6.1 Sol on 17 Oct 2026), so it includes forecasts from the versions active on each date; the current model name identifies the line, not every historical forecast.

Calibration

Stated confidence against realised accuracy. The pooled all-model row of the same table is printed next to GPT-6.1 Sol's, because a gap only means something against the population it was measured in.

ScopenHit rateMean stated conf.GapBrier95% Wilson
Weekly Calibration #15GPT-6.1 Sol1,03451.1%64.6%+13.5pp0.2719[48.0%, 54.1%]
Weekly Calibration #15field, all models5,66948.5%60.9%+12.4pp0.2677—
Monthly Calibration #3GPT-6.1 Sol3,62748.9%65.8%+16.9pp0.2816[47.3%, 50.5%]
Monthly Calibration #3field, all models22,57948.6%61.8%+13.2pp0.2701—

Weekly Calibration #15 · Nov 9-15, 2026 UTC · cutoff Mon 16 Nov 2026 16:00 UTC — Monthly Calibration #3 · Oct 1-31, 2026 UTC · cutoff Sun 1 Nov 2026 16:00 UTC

Model-line row — version change in this window: ChatGPT 5.6 Sol → GPT-6.1 Sol on 17 Oct 2026 (see the note above).

Gap vs prior window. 17.4pp → 13.5pp (-3.9pp), as Weekly Calibration #15 printed it.
Stated confidence bucketnHit rateMean stated conf.
50-6012748.8%58.3%
60-7076352.8%64.2%
70-8014443.8%71.9%
80-1000no datano data

Buckets as Weekly Calibration #15 printed them; a cell the issue marked insufficient or empty keeps its count and prints no rate.

Stability — ChatGPT 5.6 Sol, Run 2

Repeated responses from ChatGPT 5.6 Sol to identical payloads. These historical results belong to the predecessor of GPT-6.1 Sol. The caption on each tile is the same measurement in the comparison run shipped with the issue.

Unanimity
88.9%
baseline (Jul 31, as ChatGPT 5.6 Sol): 70.4%
Hard flips
0
baseline (Jul 31, as ChatGPT 5.6 Sol): 0
Soft flips
3
baseline (Jul 31, as ChatGPT 5.6 Sol): 8
Invalid
0.0%
baseline (Jul 31, as ChatGPT 5.6 Sol): 0.0%
Sideways
62.2%
baseline (Jul 31, as ChatGPT 5.6 Sol): 15.6%
Confidence SD
1.2pp
baseline (Jul 31, as ChatGPT 5.6 Sol): 2.3pp

Stability Index — Run 2 (as ChatGPT 5.6 Sol) · 10 Aug 2026 · 945 calls, 5 repeats per set, 27 sets per model, 3 tickers · compared against baseline (Jul 31, as ChatGPT 5.6 Sol)

By horizon

GPT-6.1 Sol split by forecast horizon, as the Calibration issues print it. Only the horizons inside the reports' scoring gate appear here; a cell the builder annotated keeps its annotation.

HorizonWeekly nWeekly hit rateMonthly nMonthly hit rate
1H59751.9%2,18949.8%
4H36746.6%1,21347.2%
1D7067.1%22549.8%

Weekly Calibration #15 · Nov 9-15, 2026 UTC · cutoff Mon 16 Nov 2026 16:00 UTC — Monthly Calibration #3 · Oct 1-31, 2026 UTC · cutoff Sun 1 Nov 2026 16:00 UTC

Model-line row — version change in this window: ChatGPT 5.6 Sol → GPT-6.1 Sol on 17 Oct 2026 (see the note above).

By asset

The Model Watch issues print a best-pairs and a worst-pairs table — the model-instrument cells that stood out in the window. Only GPT-6.1 Sol's own rows are listed; a model that made neither list simply has no row there.

InstrumentListnHit rateIssue
SOLbest pairs21455.6%Weekly Model Watch #15
BNBbest pairs18851.6%Weekly Model Watch #15
BTCbest pairs75650.1%Monthly Model Watch #3

Weekly Model Watch #15 · Nov 9-15, 2026 UTC · cutoff Mon 16 Nov 2026 16:00 UTC — Monthly Model Watch #3 · Oct 1-31, 2026 UTC · cutoff Sun 1 Nov 2026 16:00 UTC

Model-line row — version change in this window: ChatGPT 5.6 Sol → GPT-6.1 Sol on 17 Oct 2026 (see the note above).

Long
6.1%
share of the model's calls
Short
71.7%
share of the model's calls
Sideways
22.3%
share of the model's calls
Long / short
0.08
ratio, not a share
Mature calls
1,330
behind the mix

Weekly Model Watch #15 · Nov 9-15, 2026 UTC · cutoff Mon 16 Nov 2026 16:00 UTC

Consensus

Forecasts are grouped by whether their direction matched or opposed the leave-one-out majority of the other eligible model lines. GPT-6.1 Sol is scored separately on each group.

ScopeAgree nAgree hit rateDisagree nDisagree hit rateLiftThin or tied
Consensus Watch #1578249.5%2070.0%-20.5pp232
Monthly Consensus Watch #32,97049.3%2924.1%+25.1pp628

leave-one-out: matched vs opposed the majority of the other model lines · Consensus Watch #15 · Nov 9-15, 2026 UTC · cutoff Mon 16 Nov 2026 16:00 UTC — Monthly Consensus Watch #3 · Oct 1-31, 2026 UTC · cutoff Sun 1 Nov 2026 16:00 UTC

Model-line row — version change in this window: ChatGPT 5.6 Sol → GPT-6.1 Sol on 17 Oct 2026 (see the note above).

Paper trading (scored separately from prediction quality)

What a mechanical paper-trading rule made of GPT-6.1 Sol's forecasts in the same windows. These figures are kept apart from the accuracy and calibration figures above: a model can be well calibrated and unprofitable, or profitable and badly calibrated.

ScopeTradesWin rateNet PnLGross PnLMax drawdown
Weekly Calibration #151,03446.4%$41.66$145.06-$79.41
Monthly Calibration #33,62739.5%-$280.19$82.51-$288.45

Weekly Calibration #15 · Nov 9-15, 2026 UTC · cutoff Mon 16 Nov 2026 16:00 UTC — Monthly Calibration #3 · Oct 1-31, 2026 UTC · cutoff Sun 1 Nov 2026 16:00 UTC

Model-line row — version change in this window: ChatGPT 5.6 Sol → GPT-6.1 Sol on 17 Oct 2026 (see the note above).

Published history

Every window GPT-6.1 Sol's slot has been published in, oldest first. Model Watch supplies the count, the hit rate, the interval and the rank; Calibration adds Brier, the gap and the mean stated confidence. The cadences are separate tables because a quarter, a month and a week are different windows over the same forecasts, and a window an older issue also filed under a superseded id keeps that reading on its own line: no issue published the two as one figure.

Weekly windows

15 published windows, from Aug 3-9, 2026 UTC to Nov 9-15, 2026 UTC
Issue (model line)WindownHit rate95% WilsonRankBrierGapMean stated conf.
#1 · Model Watch · Calibrationgpt-5.6-solearlier versionAug 3-9, 2026 UTC61243.0%[39.1%, 46.9%]#30.3076+24.4pp67.4%
#2 · Model Watch · Calibrationgpt-5.6-solearlier versionAug 10-16, 2026 UTC63644.2%[40.4%, 48.1%]#20.3090+24.0pp68.2%
#3 · Model Watch · Calibrationgpt-5.6-solearlier versionAug 17-23, 2026 UTC68853.6%[49.9%, 57.3%]#40.2710+13.9pp67.5%
#4 · Model Watch · Calibrationgpt-5.6-solearlier versionAug 24-30, 2026 UTC72942.7%[39.1%, 46.3%]#40.3133+24.8pp67.5%
#5 · Model Watch · Calibrationgpt-5.6-solearlier versionAug 31-Sep 6, 2026 UTC68941.6%[38.0%, 45.4%]#60.3183+26.2pp67.8%
#6 · Model Watch · Calibrationgpt-5.6-solearlier versionSep 7-13, 2026 UTC70940.1%[36.5%, 43.7%]#30.3230+28.4pp68.5%
#7 · Model Watch · Calibrationgpt-5.6-solearlier versionSep 14-20, 2026 UTC82149.7%[46.3%, 53.1%]#50.2876+18.4pp68.1%
#8 · Model Watch · Calibrationgpt-5.6-solearlier versionSep 21-27, 2026 UTC74446.8%[43.2%, 50.4%]#40.2951+21.0pp67.7%
#9 · Model Watch · Calibrationgpt-5.6-solearlier versionSep 28-Oct 4, 2026 UTC65146.5%[42.7%, 50.4%]#70.2951+21.3pp67.9%
#10 · Model Watch · Calibrationgpt-5.6-solearlier versionOct 5-11, 2026 UTC73443.5%[39.9%, 47.1%]#30.3059+24.7pp68.1%
#11 · Model Watch · CalibrationOct 12-18, 2026 UTC81546.5%[43.1%, 49.9%]#30.2971+21.5pp68.0%
#12 · Model Watch · CalibrationOct 19-25, 2026 UTC96257.6%[54.4%, 60.7%]#30.2468+5.6pp63.2%
#13 · Model Watch · CalibrationOct 26-Nov 1, 2026 UTC82942.5%[39.1%, 45.9%]#20.2922+20.7pp63.2%
#14 · Model Watch · CalibrationNov 2-8, 2026 UTC80745.6%[42.2%, 49.0%]#20.2804+17.4pp63.0%
#15 · Model Watch · CalibrationNov 9-15, 2026 UTC1,03451.1%[48.0%, 54.1%]#10.2719+13.5pp64.6%

Model-line row — version change in this window: ChatGPT 5.6 Sol → GPT-6.1 Sol on 17 Oct 2026 (see the note above).

Monthly windows

3 published windows, from Aug 1-31, 2026 UTC to Oct 1-31, 2026 UTC
Issue (model line)WindownHit rate95% WilsonRankBrierGapMean stated conf.
#1 · Model Watch · Calibrationgpt-5.6-solearlier versionAug 1-31, 2026 UTC2,97045.9%[44.1%, 47.6%]#40.3005+21.8pp67.7%
#2 · Model Watch · Calibrationgpt-5.6-solearlier versionSep 1-30, 2026 UTC3,13344.5%[42.7%, 46.2%]#60.3054+23.5pp68.0%
#3 · Model Watch · CalibrationOct 1-31, 2026 UTC3,62748.9%[47.3%, 50.5%]#20.2816+16.9pp65.8%

Model-line row — version change in this window: ChatGPT 5.6 Sol → GPT-6.1 Sol on 17 Oct 2026 (see the note above).

Quarterly windows

1 published window: Jul 15-Sep 30, 2026 UTC
IssueWindownHit rate95% WilsonRankBrierGapMean stated conf.
#1 · Model Watch · Calibrationgpt-5.6-solearlier versionJul 15-Sep 30, 2026 UTC9,18844.9%[43.9%, 45.9%]#70.3014+22.5pp67.4%

Each figure is the one its issue froze.

Definitions & method

Quoted from the issues these figures come from.

  • Hit. direction: exit tp1/tp2 -> hit, sl -> miss, expiry -> sign of gross pnl
  • Reading the hit rate. Hit rate follows the benchmark's TP/SL/expiry outcome rules; it is not simply the asset's price direction at the end of the forecast horizon. The weekly and monthly metrics shown here use the 1h, 4h and 1d horizons; stability runs have their own test coverage.
  • Scored pool. directional forecasts resolved inside the FH gate (1h/4h/1d) since Jul 11, at the issue cutoff
  • Gap and Brier. Gap pp = mean stated confidence minus hit-rate, in percentage points; positive means overconfident.
  • Coverage. calls scored divided by the mature calls available to that model
  • Confidence buckets. Models report confidence at discrete levels, not a continuous scale; the buckets reflect those natural breakpoints. Cells below N=10 are marked insufficient.
  • Wilson intervals. 95% Wilson CIs shown are descriptive, not inferential: observations inside one window are dependent, so read them as a range, not a formal coverage guarantee.
  • Weekly title. >=5pp gap AND non-overlapping 95% CI, N>=10
  • Paper trading. A mechanical simulation on Binance USDT-M futures, scored separately from prediction quality and reported on its own. The starting bank, the per-trade notional and the taker fee are set out on Benchmarks methodology.

Methodology v1.1 (2026-08-10) · hash e66c7e8c864a2233 · Methodology · Benchmarks methodology · Dataset