TL;DR
Luna had much lower historical cost than Claude Opus 5 in the August 8, 2026 planning replay: $0.0012 versus $0.3159 per decision, with median latency of 8.6 versus 26.6 seconds. The 29-conversation paired quality comparison was inconclusive. Those cost and timing observations support Luna for this router, without establishing a general intelligence advantage or quoting current API prices.
Planning models on 29 frozen conversations: published results
Tick: recorded mean score. Whisker: mean ± one recorded standard error (SE), not a confidence interval. Overlap alone does not establish a ranking.
16-model campaign, 29 frozen conversations; 14 model rows shown in the public table. Published 2026-08-08.
Show the table · 2 rows, 5 columns
Scroll sideways to compare all columns. Select a heading to sort.
| openai/gpt-5.6-luna | 0.731 | 0.068 | $0.0012 | 8.6s |
| anthropic/claude-opus-5 | 0.662 | 0.074 | $0.3159 | 26.6s |
No matching rows. Clear the filter to see all records.
16-model campaign, 29 frozen conversations; 14 model rows shown in the public table. Published 2026-08-08.
Which fits the tested job?
For this router, Luna offers a much lower recorded bill without a demonstrated quality loss against Opus. The published paired quality comparison was inconclusive; the cost and latency difference, rather than a proven intelligence advantage, supports the Luna choice.
Planning models on 29 frozen conversations
Luna's mean was 0.731 (SE 0.068) versus Opus's 0.662 (SE 0.074). Billed cost was $0.0012 versus $0.3159 per decision; p50 was 8.6s versus 26.6s. Use those recorded operating costs for this replay, and keep the source's inconclusive paired comparison visible. They are not current API price quotes.
Route a request and plan the work: follow the published task
A useful planner chooses the right kind of work before doing it. The complete customer conversations and system prompts remain private.
Published task summary
An incoming request and its existing conversation.
Choose whether to write a plan, call a catalog tool, ask for clarification or continue an existing agent.
Read what came back
A routing decision and plan. Every model receives the same saved conversation; the test compares that decision before fresh research begins.
Replay 29 frozen conversations across the model arms, then compare graded plan scores, billing and response time.
Results from tasks both models attempted
The tested model IDs are openai/gpt-5.6-luna and anthropic/claude-opus-5. Every result table below comes from a study in which both participated.
Compare the measured quality, historical billed cost and elapsed time within the same task and rubric.
Scores from different tasks are not combined into one winner score.
Planning models on 29 frozen conversations: read these measurements
Read the complete rubric, sample and billing notes
16-model campaign, 29 frozen conversations; 14 model rows shown in the public table
Route an incoming request: write a plan, call a catalog tool, ask for clarification or continue an existing agent.
Scoring
Top planning means were close at this sample size. The source's paired comparison with Opus was inconclusive.
Cost units
Cost is provider-billed dollars per decision. Prefix reuse can make it differ from list price multiplied by every input token.
Time units
p50 is median latency in seconds, not the duration of complete research.
Frozen conversation
A saved input kept identical across model arms so their decisions can be compared.
Standard error (SE)
An estimate of uncertainty around the sample mean. A small mean difference needs more evidence than the ranking alone.
No comments yet.