name = "Qwen3.7 Max"
description = "Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks"
family = "qwen"
release_date = "2026-05-21"
last_updated = "2026-05-21"
attachment = false
reasoning = true
temperature = true
tool_call = true
open_weights = false

[limit]
context = 1_000_000
output = 65_536

[modalities]
input = ["text"]
output = ["text"]

[[benchmarks]]
name = "SWE-Bench Verified"
score = 80.4
metric = "resolved"
source = "https://qwen.ai/blog?id=qwen3.7"
date = "2026-05-19"

[[benchmarks]]
name = "SWE-Bench Pro"
score = 60.6
metric = "resolve rate"
source = "https://qwen.ai/blog?id=qwen3.7"
date = "2026-05-19"

[[benchmarks]]
name = "SWE-Bench Multilingual"
score = 78.3
metric = "resolve rate"
source = "https://qwen.ai/blog?id=qwen3.7"
date = "2026-05-19"

[[benchmarks]]
name = "Terminal-Bench"
score = 69.7
metric = "success rate"
harness = "Terminus-2"
version = "2.0"
source = "https://qwen.ai/blog?id=qwen3.7"
date = "2026-05-19"

[[benchmarks]]
name = "GPQA Diamond"
score = 92.4
metric = "accuracy"
source = "https://qwen.ai/blog?id=qwen3.7"
date = "2026-05-19"

[[benchmarks]]
name = "Humanity's Last Exam"
score = 41.4
metric = "accuracy"
source = "https://qwen.ai/blog?id=qwen3.7"
date = "2026-05-19"

[[benchmarks]]
name = "SciCode"
score = 53.5
source = "https://qwen.ai/blog?id=qwen3.7"
date = "2026-05-19"

[[benchmarks]]
name = "MCP Atlas"
score = 76.4
metric = "success rate"
source = "https://qwen.ai/blog?id=qwen3.7"
date = "2026-05-19"

[[benchmarks]]
name = "NL2Repo"
score = 47.2
harness = "Claude Code"
source = "https://qwen.ai/blog?id=qwen3.7"
date = "2026-05-19"