OpenCode Go — model evals & what each model is for

Every model currently served by OpenCode Go: a task-by-task performance matrix, cost/value analysis, and plain-language notes on what each model is actually targeted for — agentic work, coding, data analysis, image understanding, research, or high-volume chores.

Task matrix

0
100 White outline = best model for that task · click a column header to sort · hover a cell for the source benchmark

Capability vs. price

Overall index vs. blended Go price
x = blended $/1M tokens (70% input / 30% output), y = Artificial Analysis Intelligence Index. Up and to the left is better.
Value: Intelligence Index per dollar
Index points per blended $1M tokens. Free models are excluded from the ranking but listed underneath.

What each model is for

Short answers, written for people choosing a model for a job rather than reading a leaderboard. Filter by the kind of work you need.