New: Qwen4-Max eval report is here.Read the report ↗
Updated: Data as of:

Compare models

Find the right model, side by side.

First model
Second model

Model specifications

Specifications for Qwen4-Max and GPT-6 Astra. Prices and performance use the selected provider offering.
SpecificationsQwen4-MaxQwenGPT-6 AstraOpenAI
Provider
Context sizetokens1,000,0001,000,000
Cutoff dateNot reported2026-04-30
Input costUSD / 1M tokens$1.50$10.00
Output costUSD / 1M tokens$4.50$50.00
Max outputtokens262,144128,000
LatencyTTFT · secondsNot reported365.127
Speedtokens / secNot reported70.564

Blended token cost

USD per 1M tokens · 50% input + 50% output · lower is better

Blended token cost. Qwen4-Max: $3.00; GPT-6 Astra: $30.00$3.00Qwen4-Max$30.00GPT-6 Astra

Output speed by effort

End-to-end output tokens per second · higher is better

Scroll to see all efforts →
Output speed by effort. Qwen4-Max: Not reported; GPT-6 Astra (low): 40.2; GPT-6 Astra (medium): 43.4; GPT-6 Astra (high): 39.6; GPT-6 Astra (xhigh): 39.2; GPT-6 Astra (max): 37.5Not reportedUnreported40.2Low43.4Medium39.6High39.2Xhigh37.5Max

First-party APIs. Five-day workload averages, checked November 10, 2026. Methodology ↗

  • Qwen4-Max: No effort-level measurements available.

Latency by effort

Seconds until the first answer, including thinking · lower is better

Scroll to see all efforts →
Latency by effort. Qwen4-Max: Not reported; GPT-6 Astra (low): 3.29; GPT-6 Astra (medium): 5.69; GPT-6 Astra (high): 61.1; GPT-6 Astra (xhigh): 107.7; GPT-6 Astra (max): 239.21Not reportedUnreported3.29Low5.69Medium61.1High107.7Xhigh239.21Max

First-party APIs. Five-day workload averages, checked November 10, 2026. Methodology ↗

  • Qwen4-Max: No effort-level measurements available.

Benchmark performance

5 shared benchmark comparisons. Versions and setups matched.

Qwen4-MaxGPT-6 Astra

Agent Index

Composite of 7 benchmarks · score out of 100 · higher is better

View index ↗
Agent Index. Qwen4-Max: Not reported; GPT-6 Astra: 86.2Not reportedQwen4-Max86.2GPT-6 Astra
  • Qwen4-Max: Needs all 7 benchmark results (4/7 available).
Detailed comparison reports

Practical guidance on model selection, beyond the quick benchmark tables.

Qwen4-Max or Claude Fable 5.2? ↗