docketrouter
Models / Qwen

Qwen3.7 Flash

by Qwen · qwen/qwen3.7-flash

Qwen3.7 Flash is a vision-language reasoning model from Alibaba. It is suited for multimodal agents, visual coding, search, and computer interaction, with strengths in object recognition, spatial understanding, and real-world...

reasoningtool-usevisionreleased 2026-07-27
Legal score · raw
-
not yet benchmarked
Context
1M
max output 66K
Input
$0.03
per 1M tokens
Output
$0.13
per 1M tokens
Suite cost
-
run the suite to see

Benchmark results

TaskCategoryRawJuicedCorrectLatencyCostRan
Hearsay IdentificationEvidence------
Bluebook Citation FormatResearch & Writing------
Federal Civil ProcedureProcedure------
Limitations ArithmeticProcedure------
Contract Clause ClassificationContracts------
Citation Hallucination ResistanceReliability------

Measured by DocketBuster

These numbers come from DocketBuster's own legal battery, not from DocketRouter's suite. Latest run per metric, with n and a 95% Wilson interval where the source reports one. See docketbuster.com/benchmarks.

MetricValuenIntervalMeasured
Statute pinpoint, exact section (no retrieval)8.0% (24/300)30095% CI 5.4% to 11.6%2026-08-22
Statute pinpoint, exact section (with DocketBuster retrieval)9.1% (28/307)30795% CI 6.4% to 12.9%2026-08-24
Say-nothing rate (declines to bluff when the answer is not in the record)100.0%29195% CI 98.7% to 100.0%2026-08-22
Abstained on statute pinpoint90.6% (278/307)307count, no interval reported2026-08-24
Coaching quality (GW-14x, 0 to 8)7.28 / 8-rubric mean, no interval reported2026-08-23

Source files: hard-llm-ortier-qwen37flash.json, gw14x-ortier-qwen37flash.json, hard-llm-level-qwen37flash-statute_rag.json. Raw model name in source: qwen/qwen3.7-flash.