4 models
Ranked by Terminal-Bench Hard
Reasoning Tools JSON Open weights 35.6

Newer StepFun flash model for faster agents, coding, and multimodal prompts

stepfun/step-3.7-flash 2026-05-29 256K context $0.185/M input $1.11/M output
17 providers
Open weights 27.3

StepFun flash lane for quick multimodal reasoning and coding assistance

stepfun/step-3.5-flash 2026-01-29 256K context $0.1/M input $0.3/M output
14 providers
3.8

Mistral model for multilingual chat, reasoning, and tool-assisted workflows

mistral/mistral-medium-2505 2025-05-07 131.072K context $0.4/M input $2/M output
10 providers
Tools Open weights 3.0

Popular open Llama workhorse for multilingual chat, coding, and self-hosting

meta/llama-3.3-70b-instruct 2024-12-06 128K context $0.1/M input $0.32/M output
25 providers