diff --git a/server/sched.go b/server/sched.go
index fa034d28..1102a690 100644
--- a/server/sched.go
+++ b/server/sched.go
@@ -46,6 +46,7 @@ type Scheduler struct {
 // TODO set this to zero after a release or two, to enable multiple models by default
 var loadedMax = 1          // Maximum runners; < 1 maps to as many as will fit in VRAM (unlimited for CPU runners)
 var maxQueuedRequests = 10 // TODO configurable
+var numParallel = 1
 
 func InitScheduler(ctx context.Context) *Scheduler {
 	maxRunners := os.Getenv("OLLAMA_MAX_LOADED_MODELS")
@@ -57,6 +58,14 @@ func InitScheduler(ctx context.Context) *Scheduler {
 			loadedMax = m
 		}
 	}
+	if onp := os.Getenv("OLLAMA_NUM_PARALLEL"); onp != "" {
+		p, err := strconv.Atoi(onp)
+		if err != nil || p <= 0 {
+			slog.Error("invalid parallel setting, must be greater than zero", "OLLAMA_NUM_PARALLEL", onp, "error", err)
+		} else {
+			numParallel = p
+		}
+	}
 
 	sched := &Scheduler{
 		pendingReqCh:  make(chan *LlmRequest, maxQueuedRequests),
@@ -81,6 +90,8 @@ func (s *Scheduler) GetRunner(c context.Context, model *Model, opts api.Options,
 		successCh:       make(chan *runnerRef),
 		errCh:           make(chan error, 1),
 	}
+	// context split across parallel threads
+	opts.NumCtx = opts.NumCtx * numParallel
 	select {
 	case s.pendingReqCh <- req:
 	default: