Back to model catalog

Compare models

Context, limits, capabilities, price and use cases side by side — up to 4 models at a time. The URL holds your selection, so you can share it.

Qwen4 MaxQwen3.8 MaxChat with these 2
AttributeQwen4 Maxqwen/qwen4-maxQwen3.8 Maxqwen/qwen3.8-max
Specs
Context window1M1M
Max output393K131K
Input modalitiesTextImageVideoTextImageVideo
Output modalitiesTextText
Capabilities
StreamingSupportedSupported
Tool callingSupportedSupported
Structured output (JSON)SupportedSupported
Image inputSupportedSupported
Thinking modeSupportedSupported
Price / 1M tokens
Input$ 1.60$ 1.88
Output$ 4.80$ 5.63
Cache read$ 0.080$ 0.23
Cache write$ 2.00$ 2.35
What it is good at
Best for
  • Professional and industry work
  • Long-horizon agents
  • Million-token context
  • Maximum reasoning
  • Multimodal coding agents
Published benchmarks
Terminal-Bench 2.1pass_rate—
  • 86.6%Claude Code · 10 runs(setup not fully disclosed)
Terminal-Bench 4.0pass_rate
  • 48.5%(setup not fully disclosed)
—
DeepSWE 1.1resolved_rate
  • 75.7%(setup not fully disclosed)
—
ProgramBenchscore
  • 31.5%(setup not fully disclosed)
—
SWE-Bench Proresolved_rate—
  • 67.7%Claude Code(setup not fully disclosed)
OSWorld Verifiedsuccess_rate—
  • 86.1%(setup not fully disclosed)
HealthBench Professionalscore
  • 60%(setup not fully disclosed)
—
MRCR v2 256K (8-needle)score—
  • 92.9%(setup not fully disclosed)
Chartography (no tools)score
  • 54.2%(setup not fully disclosed)
—
MMMU Proaccuracy—
  • 82.3%(setup not fully disclosed)
Humanity's Last Examaccuracy
  • 51.7%(setup not fully disclosed)
  • 43.6%(setup not fully disclosed)
PaperBench (BasicAgent)score—
  • 93%3 runs(setup not fully disclosed)
GPQA Diamondaccuracy—
  • 92.6%(setup not fully disclosed)
Toolathlon Verifiedpass_rate—
  • 72.5%(setup not fully disclosed)
VideoMMMUaccuracy—
  • 88.7%(setup not fully disclosed)