OpenCode
An open-source AI coding agent for the terminal.
- Type
- Builders
- Last shipped
- Oct 21, 2026
Everything OpenCode shipped
October 2026
- New model
OpenCode added Meta Muse Wonder as a model users can select.
- New model
OpenCode revealed its stealth model Space Bunny as MiniMax M3.1 Flash and made it available, free for the next 3 days.
- New model
OpenCode updated DeepSeek V4.1 Flash to the 1017 version for users of OpenCode Go and the free tier.
- Pricing
OpenCode made DeepSeek V4.1 Flash free to use.
- New model
OpenCode added NVIDIA's Nemotron 3.5 Super VL, a 120B multimodal model built for long-running agents, free to use.
- New model
OpenCode added the stealth model Cat Jam, free to use, with zero data retention and a focus on agentic coding.
- New model
OpenCode added Gemini 4 Argon as an available model.
- New model
OpenCode made Step 5 Preview free for a week, with 1M context, multimodal input and zero data retention.
- New model
OpenCode added Claude 5.5 Haiku for OpenCode and OpenCode Go users.
- New model
OpenCode added Mistral Large 4, priced 50% off for the first two weeks.
- New model
OpenCode made inclusionAI's Ling-3.1-flash free to use, a 560B model with a 262K context window.
September 2026
- New model
OpenCode added GPT 6.1 Sol as an available model.
- Pricing
OpenCode launched Go Plus, a $40 per month plan with higher usage limits.
- Pricing
OpenCode made its $60 of usage for DeepSeek V4.1 Flash a permanent offer.
- New model
OpenCode made the stealth model Space Bunny free for a week, with 1M context, multimodal input and zero data retention.
- New model
OpenCode added GPT 6 Sol, GPT 6 Luna and Opus 5.5 as available models.
- Pricing
OpenCode cut the price of Grok 4.7 by 30% for one week.
- Pricing
OpenCode made MiMo V2.6 Flash free for a week and added both MiMo Flash and Pro to OpenCode Go.
- New model
OpenCode made the stealth model Union Alpha free for a week, with image support and no data training.
- New model
OpenCode added DeepSeek Flash v4.1 to OpenCode Go.
- Pricing
OpenCode doubled the usage limits for GLM 5.3 Flash on OpenCode Go.
- Pricing
OpenCode raised the rate limits on Meta Muse Spark Contributor to reduce 429 errors.
- New model
OpenCode added Omen Alpha, a new stealth model, exclusively for OpenCode Go subscribers, with $100 of usage for $10.
- New feature
OpenCode Go will start returning errors on 09/06 for requests missing the x-opencode-session header, which tools need to send for prompt caching.
- New model
OpenCode added Fable 5.1 as an available model.
- New model
OpenCode added Meta Muse Spark 1.3 Contributor to OpenCode Go for subscribers.
August 2026
- New model
OpenCode added the Hy4 preview model to OpenCode Go, with a 1M context window built for coding agents.
- New model
OpenCode added GLM-5.3-Flash, formerly Ox Alpha, to OpenCode Go with double usage for a limited time.
- New model
OpenCode added Grok 4.6 to OpenCode Go, with 169 requests every 5 hours.
- New model
OpenCode added the fully open source LongCat-2.0 model, with a 1M context window, to OpenCode Go.