Beating AI News Flash: OpenCode announces limited-time free access to DeepSeek V4.1 Flash. Users can now select "DeepSeek V4.1 Flash Free" in the model list, with no charges for both input and output tokens. However, the free version only offers a context window of 200,000 tokens, compared to 1 million for the paid version—a reduction of 80%. When handling large code repositories or running Coding Agents for extended periods, the free version is more likely to hit the context limit.
DeepSeek V4.1 Flash was integrated into OpenCode as early as September 10. On October 12, OpenCode had just increased the V4.1 Flash usage quota in its $10-per-month Go plan to a value of $80, and now it has added a separate free tier.
OpenCode's official documentation confirms that this free access is time-limited, but no end date has been announced yet. The platform also promises that the free-tier service provider follows a zero data retention policy and will not use user content for model training.

