Kimi K2.6

AA reports Kimi K2.6 at 54 and notes 170M tokens generated during evaluation; Kilo lists 54.4% coding completion at $24.84 per attempt.

70 Floor score

confidence: medium

70 Floor Capability

Finished benchmark shape.

77 Operating Envelope

Cheapest tracked DSWE row, but slow and verbose enough to cap the score.

76 Frontier Reality

AA Index 54 / Kilo #6 coding

66 Research Trust

Strong open/current model, but token volume and verbosity make trust scoring cautious.

Current signal

AA Index 54 / Kilo #6 coding

Evidence completeness

64%

5 score dimensions represented.

DSWE

24% pass@1

$3.16 / 56m / 84k output tokens

Calculation

Score components

70 Accuracy + grounding

Correct, grounded answers on ordinary verifiable tasks.

70 Instruction discipline

Follows exact user constraints and output formats.

66 Honesty + reliability

Avoids unsupported confidence and remains stable across runs.

72 Reasoning floor

Handles non-flashy reasoning without brittle failures.

74 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

Strong open/current model. AA notes very verbose token generation, so operating metrics must be captured on every internal run.

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 Kilo Bench Kimi release notes