History

Latest AIQ snapshot

Compare every published configuration now; trend lines begin after the next Official cycle.

Published evidence

History: Values come from RLS-protected public reads.

Measure
Family
ViewSnapshot

First published observation. Compare the configurations in this snapshot; trend lines begin after the next published cycle.

Interactive chart. The complete values are available in the following data table.

17 published series · scoring 1.0.8 · first observation; trend begins with the next published cycle

Latest visible date: 2026-08-28 UTC · 17 observations · calibrated ability range 69.477.6. Highest point estimate: sol-medium · conditional 95% interval 65.0–86.6 · strict pass 27.8% (n=72) · n=72 · coverage Unavailable · scoring 1.0.8 · published

Evidence notes and visible values

Showing 17 configurations in canonical matrix order. Family and reasoning are explicit filters, not point-estimate cutoffs. Published buckets retain the latest exact Official run; synthetic fixture buckets expose no run detail. Lines connect observations for continuity only; they do not interpolate or estimate values between dates. Absent buckets remain gaps, bars use a zero baseline, and the server returns at most 20 buckets per configuration. Each point requires matching run, configuration, scoring version, and provenance identity. Scoring versions: 1.0.8.

RecordedSeriesPrimary metricPrimary intervalStrict passnRun / bucketCoverageRuntimeMissingSummed adapter durationAPI-equivalent costScoringEvidence
2026-08-28sol-low76.4 · Calibrated ability63.5–85.7 · Conditional 95% interval26.4% (n=72)72run_602248d32fb259e607d90a9fc14707df48e308f4ede3921cc6324341cde06900
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28sol-medium77.6 · Calibrated ability65.0–86.6 · Conditional 95% interval27.8% (n=72)72run_5a2a4eca1cb69ceb798ada0d14415156ac4ab27f458ab646760fecac602b31b1
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28sol-high73.2 · Calibrated ability59.9–83.4 · Conditional 95% interval25.0% (n=72)72run_437b45de053bfca2ff1f69ad7dc57c006818f1ab4203786cc7db5cb6cde076ac
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28sol-xhigh75.9 · Calibrated ability63.0–85.4 · Conditional 95% interval27.8% (n=72)72run_b2372624b77abd0598e1c4f3b783a3726ec32241eacc3396d803f4ae0641b430
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28sol-max75.2 · Calibrated ability62.2–84.9 · Conditional 95% interval26.4% (n=72)72run_366299d303a738a36a1289a57a3234cc7059d0b057e620f6139ac02787899be3
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28sol-ultra75.2 · Calibrated ability62.2–84.9 · Conditional 95% interval27.8% (n=72)72run_395ab5e170b1b9653667e306f6c7ffd93d118175c19adacb0a7c5d5b8bfca7b1
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28terra-low70.0 · Calibrated ability56.2–80.9 · Conditional 95% interval29.2% (n=72)72run_0333c6971437abbd8e524383310d57e9441e80de282d6cd0ba85a4183602d1be
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28terra-medium74.5 · Calibrated ability61.3–84.3 · Conditional 95% interval27.8% (n=72)72run_8832ca59be523e36b05b5516ff7281b3f711eaf10cfe3581e45b8b95d1c4992b
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28terra-high74.6 · Calibrated ability61.4–84.4 · Conditional 95% interval23.6% (n=72)72run_bde48d35c4ab92dc81c7d7b05f60687b45850990211853d6674a97b37d9c56fe
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28terra-xhigh71.6 · Calibrated ability58.0–82.1 · Conditional 95% interval26.4% (n=72)72run_81e97c23f7c4ad230af28152a1c110db8164e0a8769f2dc78ceadb049cf0d294
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28terra-max71.6 · Calibrated ability58.0–82.1 · Conditional 95% interval26.4% (n=72)72run_29b8598df64fabb999b6f034956e2a10f13758a151d32272cb8cf8746b2d0c52
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28terra-ultra72.8 · Calibrated ability59.4–83.0 · Conditional 95% interval25.0% (n=72)72run_8dc4463121678365f69e79c11278f844c6b2acdd49f76da099a46e852be079ef
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28luna-low69.4 · Calibrated ability55.6–80.4 · Conditional 95% interval25.0% (n=72)72run_bdc18132086ad72aa7b1f858785bf0673abb4507c9f6633eaad13d06e1b84ad0
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28luna-medium70.6 · Calibrated ability56.9–81.4 · Conditional 95% interval23.6% (n=72)72run_92a9f7939a2dc7b961a3cbd6b13f719907ae2adfbaca013b6f086514266747cc
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28luna-high72.7 · Calibrated ability59.2–82.9 · Conditional 95% interval27.8% (n=72)72run_9c6cefcef6a726dc2ab54d6b851f06629e45097f8de1b6d47f709ff1cccee3b4
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28luna-xhigh75.7 · Calibrated ability62.8–85.3 · Conditional 95% interval29.2% (n=72)72run_652a8005e9e3066cf1987e27a140a421e256f7961aca372489ff0686f885ed83
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
2026-08-28luna-max76.3 · Calibrated ability63.4–85.7 · Conditional 95% interval27.8% (n=72)72run_cc3cfefe0e6c96deb38f020dcb5154bda98b62333adc00bd816f9c9f6c7232ab
Latest of 1 run
UnavailableUnavailableUnavailableUnavailableUnavailable1.0.8Published
Transport measurement · not scoring

Normal vs Fast

Same fixed response task, paired modes, 5 trials per mode. AIQ is not recomputed from speed, time, tokens, or credits. TTFT is not shown because the current CLI does not expose first-token timestamps.

Median Fast speedup
1.35×
Fast wins
17/17
Unavailable modes
0
Paired configurations
17/17
Measure
View

Interactive chart. The complete values are available in the following data table.

Exact paired measurements
ConfigurationNormal completionFast completionNormal timeFast timeSpeedupNormal outputFast outputNormal creditsFast credits
sol-low100.0%100.0%21.7 s16.1 s1.35×37.4 tok/s51.3 tok/s8.103 credits20.261 credits
sol-medium100.0%100.0%22.0 s16.0 s1.38×36.5 tok/s51.4 tok/s8.150 credits18.677 credits
sol-high100.0%100.0%21.3 s16.1 s1.33×38.9 tok/s51.3 tok/s8.165 credits18.735 credits
sol-xhigh100.0%100.0%23.3 s16.6 s1.40×36.5 tok/s50.3 tok/s7.525 credits20.468 credits
sol-max100.0%100.0%22.3 s16.4 s1.36×37.9 tok/s50.9 tok/s8.229 credits18.945 credits
sol-ultra100.0%100.0%22.4 s16.2 s1.38×38.4 tok/s52.4 tok/s8.297 credits20.657 credits
terra-low80.0%100.0%24.5 s18.5 s1.32×33.7 tok/s43.3 tok/s3.130 credits6.786 credits
terra-medium100.0%100.0%29.9 s19.2 s1.56×26.8 tok/s42.3 tok/s3.240 credits6.772 credits
terra-high100.0%80.0%21.1 s16.8 s1.26×39.0 tok/s49.1 tok/s2.669 credits5.901 credits
terra-xhigh100.0%100.0%21.4 s16.2 s1.33×38.7 tok/s51.5 tok/s2.732 credits6.181 credits
terra-max100.0%100.0%21.6 s15.9 s1.36×38.8 tok/s53.0 tok/s2.491 credits5.577 credits
terra-ultra100.0%100.0%21.2 s16.7 s1.27×39.4 tok/s50.5 tok/s2.575 credits5.791 credits
luna-low100.0%100.0%21.0 s15.4 s1.36×39.4 tok/s53.0 tok/s0.291 credits0.595 credits
luna-medium100.0%100.0%21.1 s15.5 s1.36×38.9 tok/s52.7 tok/s0.269 credits0.661 credits
luna-high100.0%100.0%20.8 s16.1 s1.29×39.7 tok/s51.1 tok/s0.270 credits0.662 credits
luna-xhigh100.0%100.0%20.8 s16.2 s1.28×40.0 tok/s50.9 tok/s0.268 credits0.598 credits
luna-max100.0%100.0%21.6 s16.5 s1.30×40.1 tok/s54.4 tok/s0.297 credits0.754 credits

Aggregate output rate uses total elapsed time because the current Codex event stream does not expose a trustworthy first-token timestamp. Fast credit estimates use the published 2.5× ChatGPT credit multiplier. Neither measure changes AIQ, confidence intervals, or ranking.

Evidence availabilitypublished + unavailable · 4 sources
Published evidence

Matrix entries: Values come from RLS-protected public reads.

Published evidence

Trend points: Values come from RLS-protected public reads.

Published evidence

Historical efficiency: Values come from RLS-protected public reads.