OpenCode Go

OpenCode Go is a $10/month subscription that provides curated, low-latency access to a roster of open coding models, used through the OpenCode client like any other provider. According to the official documentation, the team tested a select group of open models, worked with providers to ensure correct serving, and benc The documented model list includes Grok 4.6, GLM-5.3-Flash, GLM-5.3, GLM-5.2, GLM-5.1, GPT 5.6 Luna, Kimi K3, Kimi K2.7 Code, Kimi K2.6, LongCat-2.0, MiMo-V2.5, MiMo-V2.5-Pro, MiniMax M3, MiniMax M2.7, Muse Spark 1.3 Contributor, Muse Spark 1.2 Contributor, Qwen3.8 Max, Qwen3.8 Flash, Qwen3.7 Max, Qwen3.7 Plus, Qwen3.6

Neowin reports that Dax Raad, co-founder of OpenCode, announced on September 12, 2026 that the company's $10-per-month OpenCode Go subscription had broken even for two consecutive weeks, calling it a major milestone given that competing offerings are either burning money or silently opting users into data-selling schem The piece lists the paid-tier models, including Kimi K3, Kimi K2.7 Code, GPT 5.6 Luna, MiniMax M3, Qwen3.7 Plus, GLM-5.3-Flash, DeepSeek V4 Flash, DeepSeek V4.1 Flash, MiMo-V2.5, and Muse Spark 1.3 Contributor, and notes that OpenCode additionally provides free models such as Nemotron 3 Ultra, Muse Spark 1.3 and 1.2, N

OpenCode Go has added Alibaba's Qwen3.8 Flash to its $10-per-month coding-model subscription, as reported by a third-party editorial on August 28, 2026. The service gives developers another managed option within the Qwen architecture, though the article notes that OpenCode does not disclose which provider serves Qwen3. The article situates the addition within OpenCode Go's broader catalog of Qwen3.8 Flash, Qwen3.8 Max, DeepSeek V4 Flash, GPT 5.6 Luna, Kimi, GLM, and MiniMax models, and explains OpenCode's curation model of testing open models, coordinating with providers on serving, and benchmarking the combination. It also reference

A KuCoin news flash reports commentary from an OpenCode engineer named dax arguing that DeepSeek's upcoming price increase is more likely a traffic-management measure than a sign of operating losses. According to the article, OpenCode claims it has already replicated DeepSeek's current pricing through partner arrangeme The piece attributes DeepSeek's caching advantage to three prefix-based disk persistence mechanisms and to V4-era Token-wise Compression combined with Sparse Attention, which it says significantly lowers the cost of million-token contexts. It closes with a forward-looking point that competition among AI coding products

Using any LLM provider in OpenCode.

On September 12, 2026, OpenCode co-founder Dax Raad announced that the $10-per-month OpenCode Go subscription has achieved break-even for two consecutive weeks, according to a secondary news report. The article frames this as a notable commercial-sustainability milestone for a low-cost open-source AI coding subscriptio The piece reiterates the $10 price point and the paid plan's model lineup, including Kimi K3, Kimi K2.7 Code, GPT 5.6 Luna, MiniMax M3, Qwen3.7 Plus, GLM-5.3-Flash, DeepSeek V4 Flash, DeepSeek V4.1 Flash, MiMo-V2.5, and Muse Spark 1.3 Contributor. It also contrasts this with OpenCode's free tier, which includes Nemotro

The OpenCode official changelog records maintenance changes that touch the Go subscription directly: outdated OpenCode Go first-month discount messaging and pricing have been removed, indicating an active cleanup of Go's promotional/positioning copy. A separate fix addresses OpenCode device-login links when servers ret Additional provider-routing fixes accompany the Go changes, including correct handling of `textVerbosity` for OpenAI-compatible providers that do not support it, and broader Cloudflare AI Gateway, Azure, Bedrock, and Vertex AI reliability improvements. These updates signal ongoing hardening of the provider layer that O