modelgrep

Anthropic: Claude Opus 4.8 vs Qwen: Qwen3.6 Plus

Anthropic: Claude Opus 4.8 wins on more metrics (5 of 8), but the right pick depends on what you optimize for — see the breakdown below.

MetricAnthropic: Claude Opus 4.8Qwen: Qwen3.6 Plus
Intelligence Index61.450.0
Coding Index56.742.9
GPQA Diamond92%88%
Design Arena Elo13721288
Speed (tokens/sec)5936
Latency1.8s1.1s
Input price /M$5.00$0.325
Output price /M$25.00$1.95
Context window1M1M
CapabilitiesReasoningToolsJSONVisionReasoningToolsJSONVision