BenchLeader
AlibabaAuto-detected

Qwen3.5 Omni Plus

Best configuration ranks #229 of 610 on the BenchLeader Index at 53.3 ±1.9. Released 30 Mar 2026.

Blended price
$1.50/M
$0.400 in · $4.80 out
Output speed
86 tok/s
First answer
2.14 s
first token 2.14 s
Context
256k
How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.
  1. Overall index53
  2. Agents & tools52
  3. Knowledge58
  4. Instruction following52
  5. Multimodal55
  6. Long context58
  7. Composite55

Benchmark results

One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.

Reasoning

BenchmarkdefaultSource
GPQA Diamond (AA)not in index82.6%#167Artificial Analysis
Humanity's Last Exam (AA)not in index14.9%#196Artificial Analysis

Agents & tools

Knowledge

BenchmarkdefaultSource
AA-Omniscience-13.1#156Artificial Analysis

Instruction following

BenchmarkdefaultSource
IFBench51.2%#166Artificial Analysis

Multimodal

BenchmarkdefaultSource
MMMU-Pro70.5%#112Artificial Analysis

Long context

BenchmarkdefaultSource
AA-LCR63.7%#202Artificial Analysis

Composite

BenchmarkdefaultSource
AA Intelligence Index20.4#170Artificial Analysis

What a task costs

Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.

WorkloadTokens in / outCostWith cachingTime
Chat reply400 / 300$0.00165.6 s
Summarise a 30-page report12,000 / 600$0.00779.1 s
Code edit6,000 / 1,500$0.009619.5 s
Agentic coding session60,000 / 4,000$0.04348.5 s
Structured extraction2,000 / 200$0.00184.5 s

See also

Data as of 9 Sept 2026. Compare with another model.