FIBRE

gpt-5.6-luna

OpenAI · gpt-5.6-luna · unpinned

Ranked 3 of 5 overall on 202 tasks. Strongest on monitoring at 91.2, weakest on credit analysis at 75.7.

Score
80.4
rank 3 of 5
Latency
24.5s
median per task
Cost
$0.013
per task at list prices

Capability profile

Mean score per category on question-answering, 0–100.
CreditMonitorStructuredResearchDiscovery

Score by category

5 categories, 152 tasks on question-answering.
Credit Analysis75.7 · 62
Monitoring91.2 · 20
Structured Products85.5 · 18
Research Intelligence76.7 · 10
Discovery & Prediction80.2 · 2

Where it sits

gpt-5.6-luna against the rest of the board. Everything else is drawn back so the one model reads.
No model here.Better on both axes at once.$0.01$0.02$0.05$0.1$0.2$0.5$165758595Cost per task · better ←Score · better ↑Best per taskgpt-5.6-luna
gpt-5.6-luna on the frontier — no model beats it on bothFrontier · 3 models nothing beats on both

Best per task is the diamond: for each of the 200 tasks every model attempted, the highest score any of them got, averaged — 95.0 against 91.3 for opus-5 used on everything. Nobody can build it. The choice is made with the result in hand, which a router does not have when it has to choose, so it marks the edge of what having 5 different models is worth rather than a score anyone can reach. It is also flattered by noise — the highest of 5 measurements rises with the count even when the models behind them do not improve — so it will drift up as models are added and is not comparable between boards with different numbers of them. 2 tasks are left out of it, for not having been attempted by every model. Its cost is measured the way the board measures a model’s, on the question-answering track that holds more of the tasks.

How it works when given tools

Behaviour across 150 agentic-research runs.
Delivered an answer
100%
of runs produced a deliverable
Mean steps
14.8
tool calls per run
Tools used per run
fetch_url 10.3sec_search 2.6web_search 1.1python_exec 0.9
How runs ended
final answer 91%max steps 9%

Where it lost the most

The lowest-scoring tasks on question-answering, with the rubric gates that fired where they are known.
PARTIAL
Relative Value
Credit Analysis · RV, technology, sbc
penalty: Does not take an extreme position — neither blindly adding back all SBC (reporting CF at…
13.0
PARTIAL
Breakeven Analysis
Credit Analysis · breakeven, carry, roll down
13.0
PARTIAL
MBS Prepayment
Structured Products · structured, MBS, dollar roll
14.3
PARTIAL
Spread Analysis
Credit Analysis · spread, accrued interest, dirty price
21.1
PARTIAL
Spread Analysis
Credit Analysis · spread, g spread, z spread
29.4