3 models
Ranked by SWE-Bench Verified
Reasoning Tools Open weights 78.9

MiMo-V2.5-Pro is Xiaomi’s flagship model, delivering strong performance in general agentic capabilities, complex software engineering, and long-horizon tasks, with top rankings on benchmarks such as ClawEval, GDPVal, and SWE-bench Pro....

xiaomi/mimo-v2.5-pro 2026-04-22 1.04858M context $0.435/M input $0.87/M output
24 providers
Reasoning Tools JSON Open weights 76.4

Large open Qwen multimodal MoE for visual agents and long technical tasks

alibaba/qwen3.5-397b-a17b 2026-02-15 262.144K context $0.6/M input $3.6/M output
18 providers
Reasoning Tools JSON Open weights 73.4

Open multimodal Qwen MoE for local agents that need vision, audio, and code

alibaba/qwen3.6-35b-a3b 2026-04-17 262.144K context $0.248/M input $1.485/M output
19 providers