Qwen3.7 Plus

OpenCode Go's official documentation confirms Qwen3.7 Plus as one of the curated open coding models available through its $10/month subscription. The docs describe the onboarding flow—signing in to OpenCode Zen, running the /connect command in the TUI, pasting the API key, and listing available models with /models—and OpenCode positions Go as a low-cost international subscription addressing the difficulty of getting reliable, low-latency access to open coding models. Qwen3.7 Plus is explicitly listed in the current model roster, giving developers a documented way to access Alibaba's multimodal agent model through OpenCode's gateway

Superframeworks reports a third-party "non-frontier" benchmark in which Qwen3.7 Plus won all three tested founder tasks—positioning/hero rewrite, pricing/packaging decision, and customer interview synthesis—against GLM 5.2, DeepSeek V4 Flash, Kimi K2.6, and MiMo V2.5 Pro. All candidates were routed through OpenCode Go The article positions the takeaway not as brand worship but as evidence that cheaper, less-celebrated models are now strong enough that founders should build a routing table for model work rather than defaulting to the most famous frontier model for every job. Qwen3.7 Plus's wins were attributed to choosing clear ICPs,

Alibaba's Qwen team released Qwen3.7-Plus on June 2, 2026, as a multimodal agent model that accepts text, images, and video as input but outputs only text. The model combines visual perception, GUI operation, and autonomous code generation in a single agent loop, enabling it to read screens, navigate applications, writ The model is generally available through Alibaba Cloud's Bailian platform (marketed internationally as Model Studio) via API, priced at $0.40 per million input tokens and $1.60 per million output tokens, roughly 6x cheaper than Qwen3.7-Max on input. On agent benchmarks, Qwen3.7-Plus leads with ScreenSpot Pro at 79.0 (v

The LLM Stats benchmark tracker ranks Qwen3.7-Plus at 43rd overall on its composite LLM Stats Score, with tier standings placing it in the "Great" category for Legal (2 of 202) and Healthcare (2 of 239), and "Good" for Chat (6 of 125) and Long Context (10 of 114). The model shows a +0.98σ improving quality signal over Per-benchmark evaluation sourced from qwen.ai and aggregator calculations places Qwen3.7-Plus at IFEval rank 2 (0.95/1), MMLU-Redux rank 3 (0.94/1), HMMT Feb 26 rank 4 (0.93/1), and MRCR v2 rank 1 (0.92/1). These independently tracked benchmark scores provide live performance corroboration for the model's capabilities

OpenRouter's listing confirms Qwen3.7 Plus as a cost-effective model in Alibaba's Qwen3.7 series, supporting text and image input with text output, priced at $0.32 per million input tokens and $1.28 per million output tokens with a 1M context window. The page notes a release date of June 3, 2026, and that the model is The listing highlights Qwen3.7 Plus's multimodal interactive hybrid agent capability: perceiving real-world scenes, reading screens, interacting with GUIs, generating code from visual references, and navigating mobile apps end-to-end. Performance metrics include a 62.58% cache hit rate, 4.27% tool call error rate, and