Qwen3.7 Max

AA reports Qwen3.7 Max at 56.6 with high verbosity during evaluation; pricing is $2.50 input / $7.50 output per 1M tokens.

74 Floor score

confidence: medium

74 Floor Capability

Finished benchmark shape.

62 Operating Envelope

Pricing is visible, but verbosity and missing DSWE rows keep the operating evidence thin.

81 Frontier Reality

AA Index 56.6

70 Research Trust

Good current Qwen signal, but verbose output raises source-discipline and citation-risk questions.

Current signal

AA Index 56.6

Evidence completeness

62%

5 score dimensions represented.

DSWE

Not reported

Keep watch for a DeepSWE row.

Calculation

Score components

72 Accuracy + grounding

Correct, grounded answers on ordinary verifiable tasks.

74 Instruction discipline

Follows exact user constraints and output formats.

70 Honesty + reliability

Avoids unsupported confidence and remains stable across runs.

78 Reasoning floor

Handles non-flashy reasoning without brittle failures.

76 Agent/coding

Completes tool, coding, and agent-style tasks when evidence exists.

68 Operating envelope

Cost, speed, token burn, verbosity, and friction.

Evidence

Sources used

Caveat

What to watch

Top current Qwen row. Verbose output is a real operating cost risk for agentic sessions.

Meaning

How to read the score

Each score is a 0-100 saturation scale. A score of 100 means that benchmark lane is complete for this model. Floor Capability is the finished benchmark shape; Operating Envelope, Frontier Reality, and Research Trust are provisional tabs that need prompt-pack backfill. Confidence is evidence quality, not another model grade. Evidence % is source coverage, not capability.

Citations

Benchmark sources

Artificial Analysis Intelligence Index Qwen3.7 release notes