diff --git a/wasm/app/app.js b/wasm/app/app.js
index 33736fc..2c61f8a 100644
--- a/wasm/app/app.js
+++ b/wasm/app/app.js
@@ -213,19 +213,28 @@ function runOnTierInWorker(tier, src, onLoading) {
     return new Promise((resolve, reject) => {
         const w = ensureWorker(tier);
         const myRunId = ++workerState.nextRunId;
+        let liveBlock = null;
         workerState.pending[tier] = { runId: myRunId, resolve, reject };
         const handler = (e) => {
             if (e.data.runId !== myRunId) return;
             if (e.data.kind === "loading") {
                 onLoading && onLoading(e.data.tier);
+            } else if (e.data.kind === "chunk") {
+                // First chunk spawns the in-progress block; subsequent
+                // chunks append to it. User sees displays land
+                // immediately, not just at the end of the run.
+                if (!liveBlock) liveBlock = startLiveBlock(tier);
+                liveBlock.append(e.data.chunk);
             } else if (e.data.kind === "done") {
                 w.removeEventListener("message", handler);
                 workerState.pending[tier] = null;
-                resolve(e.data.output);
+                resolve({ output: e.data.output, liveBlock });
             } else if (e.data.kind === "error") {
                 w.removeEventListener("message", handler);
                 workerState.pending[tier] = null;
-                reject(new Error(e.data.message));
+                const err = new Error(e.data.message);
+                err.liveBlock = liveBlock;
+                reject(err);
             }
         };
         w.addEventListener("message", handler);
@@ -233,6 +242,44 @@ function runOnTierInWorker(tier, src, onLoading) {
     });
 }
 
+// Spawn a tier-block in the output panel immediately on first streamed
+// chunk. Returns handles to append further chunks and to finalize the
+// timing header once the eval reports done. Mirrors appendBlock's
+// structure so finalized live blocks look identical to non-streamed
+// ones — same DOM, same CSS.
+function startLiveBlock(tierName) {
+    const block = document.createElement("div");
+    block.className = "tier-block";
+    const h = document.createElement("h3");
+    h.textContent = TIERS[tierName] || tierName;
+    const t = document.createElement("span");
+    t.className = "time";
+    t.textContent = "  (running…)";
+    h.appendChild(t);
+    block.appendChild(h);
+    const pre = document.createElement("pre");
+    pre.textContent = "";
+    block.appendChild(pre);
+    outputEl.appendChild(block);
+    return {
+        append(chunk) {
+            pre.textContent += chunk;
+            outputEl.scrollTop = outputEl.scrollHeight;
+        },
+        finalize(elapsedMs, kind, fullOutput) {
+            if (kind === "cancelled") t.textContent = `  (cancelled @ ${elapsedMs.toFixed(0)} ms)`;
+            else t.textContent = `  (${elapsedMs.toFixed(0)} ms)`;
+            if (kind === "error") pre.className = "err";
+            // If the streamed chunks miss anything (e.g. asm tier
+            // which doesn't stream yet), reconcile with the full
+            // output. No-op when streaming captured everything.
+            if (fullOutput && pre.textContent !== fullOutput) {
+                pre.textContent = fullOutput;
+            }
+        },
+    };
+}
+
 function cancelCurrentRun() {
     for (const t of Object.keys(workerState.workers)) {
         if (workerState.workers[t]) {
@@ -307,20 +354,26 @@ async function runAll() {
         return runOnTierInWorker(t, src, (loadingTier) => {
             setStatus(`loading ${loadingTier}…`, "busy");
         })
-            .then((output) => {
+            .then(({ output, liveBlock }) => {
                 const elapsed = performance.now() - startTimes[t];
                 delete startTimes[t];
-                appendBlock(t, elapsed, output, "ok");
+                // Streamed tiers already painted via liveBlock; just
+                // finalize the timing header. Non-streaming tiers
+                // (today: asm) get a fresh appendBlock at the end.
+                if (liveBlock) liveBlock.finalize(elapsed, "ok", output);
+                else appendBlock(t, elapsed, output, "ok");
                 return { tier: t, ok: true, elapsed };
             })
             .catch((e) => {
                 const elapsed = performance.now() - startTimes[t];
                 delete startTimes[t];
                 if (e.message === "cancelled") {
-                    appendBlock(t, elapsed, "(cancelled)", "cancelled");
+                    if (e.liveBlock) e.liveBlock.finalize(elapsed, "cancelled");
+                    else appendBlock(t, elapsed, "(cancelled)", "cancelled");
                     return { tier: t, cancelled: true, elapsed };
                 }
-                appendBlock(t, elapsed, e.message || String(e), "error");
+                if (e.liveBlock) e.liveBlock.finalize(elapsed, "error", e.message);
+                else appendBlock(t, elapsed, e.message || String(e), "error");
                 return { tier: t, error: e.message, elapsed };
             });
     });
diff --git a/wasm/app/runner.js b/wasm/app/runner.js
index aa602af..5379f6f 100644
--- a/wasm/app/runner.js
+++ b/wasm/app/runner.js
@@ -29,9 +29,9 @@ export async function getTier(name, onLoad) {
     return cache[name];
 }
 
-export async function evalOnTier(name, src, onLoad) {
+export async function evalOnTier(name, src, onLoad, onChunk) {
     const tier = await getTier(name, onLoad);
-    return tier.evalLisp(src);
+    return tier.evalLisp(src, onChunk);
 }
 
 export function heapStats(name) {
@@ -39,3 +39,20 @@ export function heapStats(name) {
     if (!tier || !tier.heapStats) return null;
     return tier.heapStats();
 }
+
+// Portal — REPL save/resume bridge. portalSave reads a snapshot blob
+// from the tier (MEMFS on the C tier, similar bridges on others as
+// they land). Returns null when the tier hasn't implemented portals
+// yet (asm-wat today) so the caller can surface a friendly message
+// instead of crashing. portalLoad is the symmetric write-in path.
+export function portalSave(name, checkpointName) {
+    const tier = cache[name];
+    if (!tier || !tier.portalSave) return null;
+    return tier.portalSave(checkpointName);
+}
+export function portalLoad(name, checkpointName, blob) {
+    const tier = cache[name];
+    if (!tier || !tier.portalLoad) return false;
+    tier.portalLoad(checkpointName, blob);
+    return true;
+}
diff --git a/wasm/app/worker.mjs b/wasm/app/worker.mjs
index 760c680..795f0eb 100644
--- a/wasm/app/worker.mjs
+++ b/wasm/app/worker.mjs
@@ -3,7 +3,7 @@
 // live ms counter actually ticks AND so cancel works (main thread
 // terminates this worker via worker.terminate()).
 
-import { evalOnTier, setBendUrl, heapStats } from "./runner.js";
+import { evalOnTier, setBendUrl, heapStats, portalSave, portalLoad } from "./runner.js";
 
 self.onmessage = async (e) => {
     const { kind } = e.data;
@@ -17,12 +17,39 @@ self.onmessage = async (e) => {
         self.postMessage({ kind: "heap", runId: e.data.runId, tier: e.data.tier, stats: heapStats(e.data.tier) });
         return;
     }
+    if (kind === "portal-save") {
+        // Pulls the most-recent /tmp/<name>.portal out of the tier's
+        // MEMFS after the lisp side ran (portal-snapshot! NAME). Main
+        // thread encrypts and stuffs it into the vault.
+        const blob = portalSave(e.data.tier, e.data.name);
+        self.postMessage({ kind: "portal-save", runId: e.data.runId, name: e.data.name, blob });
+        return;
+    }
+    if (kind === "portal-load") {
+        // Hydrates MEMFS from a vault-decrypted blob so a subsequent
+        // (portal-load! NAME) eval finds the file ready.
+        const ok = portalLoad(e.data.tier, e.data.name, e.data.blob);
+        self.postMessage({ kind: "portal-load", runId: e.data.runId, name: e.data.name, ok });
+        return;
+    }
     if (kind !== "eval") return;
     const { runId, tier, src } = e.data;
     try {
-        const output = await evalOnTier(tier, src, (loadingTier) => {
-            self.postMessage({ kind: "loading", runId, tier: loadingTier });
-        });
+        const output = await evalOnTier(
+            tier,
+            src,
+            (loadingTier) => {
+                self.postMessage({ kind: "loading", runId, tier: loadingTier });
+            },
+            // Stream every print/display from the tier to the main
+            // thread as it happens. Sync XHR inside bend!-call still
+            // blocks the worker, but displays BEFORE/AFTER the bend
+            // round-trip surface immediately instead of waiting for
+            // the whole eval to finish. Long demos feel alive.
+            (chunk) => {
+                self.postMessage({ kind: "chunk", runId, tier, chunk });
+            },
+        );
         self.postMessage({ kind: "done", runId, output });
     } catch (err) {
         self.postMessage({ kind: "error", runId, message: err && err.message ? err.message : String(err) });
diff --git a/wasm/c/lumbda-c.loader.js b/wasm/c/lumbda-c.loader.js
index 71b0e54..181d4fe 100644
--- a/wasm/c/lumbda-c.loader.js
+++ b/wasm/c/lumbda-c.loader.js
@@ -13,12 +13,25 @@ async function _bootstrap() {
     const wasmDir = new URL("./", import.meta.url).href;
     const { default: createLumbdaC } = await import(/* @vite-ignore */ factoryURL);
 
+    // Streaming output: every printf in the wasm fires Module.print
+    // synchronously. We buffer normally for the final return, but if
+    // the current eval registered an onChunk callback we also call
+    // through immediately so the worker can postMessage a chunk to
+    // the UI as each line emerges (vs the user staring at a blank
+    // panel for 18 seconds during the heavy bend call).
     let outBuf = [];
     let errBuf = [];
+    let currentOnChunk = null;
     const module = await createLumbdaC({
         locateFile: (p) => wasmDir + p,
-        print: (line) => outBuf.push(line),
-        printErr: (line) => errBuf.push(line),
+        print: (line) => {
+            outBuf.push(line);
+            if (currentOnChunk) currentOnChunk(line);
+        },
+        printErr: (line) => {
+            errBuf.push(line);
+            if (currentOnChunk) currentOnChunk(line);
+        },
     });
 
     const _init = module.cwrap("lumbda_wasm_init", null, []);
@@ -27,6 +40,12 @@ async function _bootstrap() {
 
     _init();
 
+    // /tmp must exist before (portal-snapshot! ...) writes there — the
+    // C builtin fopen("w")s the path directly so a missing dir is a
+    // hard error. Emscripten's MEMFS gives us /tmp on recent versions
+    // but make it idempotent.
+    try { module.FS.mkdir("/tmp"); } catch (e) { /* exists */ }
+
     // bend!-call host bridge — js_lumbda_bend_call inside the wasm
     // (EM_JS in lumbda_wasm_entry.c) reads globalThis._lumbdaCBendUrl
     // for the destination, then does sync XHR POST and writes the
@@ -34,10 +53,28 @@ async function _bootstrap() {
     // global so runner.js can propagate a saved playground URL.
     return {
         setBendUrl(url) { globalThis._lumbdaCBendUrl = url || null; },
-        async evalLisp(src) {
+        // portalSave / portalLoad bridge MEMFS to the JS side so the
+        // REPL can persist checkpoints to the encrypted vault and
+        // restore them on a fresh worker. Names come from the worker
+        // (REPL caller validates alphanumeric) so path traversal isn't
+        // a concern.
+        portalSave(name) {
+            const path = `/tmp/${name}.portal`;
+            try {
+                const bytes = module.FS.readFile(path);
+                return new TextDecoder().decode(bytes);
+            } catch (e) { return null; }
+        },
+        portalLoad(name, blob) {
+            const path = `/tmp/${name}.portal`;
+            module.FS.writeFile(path, blob);
+        },
+        async evalLisp(src, onChunk) {
             outBuf = [];
             errBuf = [];
+            currentOnChunk = onChunk || null;
             const errPtr = _eval(src);
+            currentOnChunk = null;
             let errMsg = "";
             if (errPtr) {
                 errMsg = module.UTF8ToString(errPtr);
diff --git a/wasm/dist-repl/c/lumbda-c.loader.js b/wasm/dist-repl/c/lumbda-c.loader.js
index 64e78c2..181d4fe 100644
--- a/wasm/dist-repl/c/lumbda-c.loader.js
+++ b/wasm/dist-repl/c/lumbda-c.loader.js
@@ -13,12 +13,25 @@ async function _bootstrap() {
     const wasmDir = new URL("./", import.meta.url).href;
     const { default: createLumbdaC } = await import(/* @vite-ignore */ factoryURL);
 
+    // Streaming output: every printf in the wasm fires Module.print
+    // synchronously. We buffer normally for the final return, but if
+    // the current eval registered an onChunk callback we also call
+    // through immediately so the worker can postMessage a chunk to
+    // the UI as each line emerges (vs the user staring at a blank
+    // panel for 18 seconds during the heavy bend call).
     let outBuf = [];
     let errBuf = [];
+    let currentOnChunk = null;
     const module = await createLumbdaC({
         locateFile: (p) => wasmDir + p,
-        print: (line) => outBuf.push(line),
-        printErr: (line) => errBuf.push(line),
+        print: (line) => {
+            outBuf.push(line);
+            if (currentOnChunk) currentOnChunk(line);
+        },
+        printErr: (line) => {
+            errBuf.push(line);
+            if (currentOnChunk) currentOnChunk(line);
+        },
     });
 
     const _init = module.cwrap("lumbda_wasm_init", null, []);
@@ -56,10 +69,12 @@ async function _bootstrap() {
             const path = `/tmp/${name}.portal`;
             module.FS.writeFile(path, blob);
         },
-        async evalLisp(src) {
+        async evalLisp(src, onChunk) {
             outBuf = [];
             errBuf = [];
+            currentOnChunk = onChunk || null;
             const errPtr = _eval(src);
+            currentOnChunk = null;
             let errMsg = "";
             if (errPtr) {
                 errMsg = module.UTF8ToString(errPtr);
diff --git a/wasm/dist-repl/python/lumbda-py.js b/wasm/dist-repl/python/lumbda-py.js
index 609cefc..34aff77 100644
--- a/wasm/dist-repl/python/lumbda-py.js
+++ b/wasm/dist-repl/python/lumbda-py.js
@@ -29,7 +29,15 @@ async function _bootstrap() {
     // the response text back. Sync XHR is the only sync HTTP available
     // in a Web Worker; perfect for the blocking eval model lumbda
     // primitives expect.
-    const refs = { bendUrl: null };
+    const refs = { bendUrl: null, currentOnChunk: null };
+    // Streaming output bridge — sys.stdout in pyodide is replaced
+    // (during _lumbda_eval) with a class whose write() calls back
+    // here. We forward to the current onChunk so the worker can
+    // postMessage chunks to the UI as work happens, instead of the
+    // user staring at a blank panel during a long bend dispatch.
+    globalThis._lumbdaPyEmitChunk = (s) => {
+        if (refs.currentOnChunk && s) refs.currentOnChunk(s);
+    };
     globalThis._lumbdaPyBendCall = (payload) => {
         if (!refs.bendUrl) return "no bend URL configured";
         try {
@@ -63,8 +71,50 @@ def _bend_call_prim(args, env):
     return str(_js_bend_call(payload))
 _env.define(lumbda.S('bend!-call'), _bend_call_prim)
 
+# Portal save/resume — REPL bridge. Mirrors the C tier's
+# (portal-snapshot! NAME) / (portal-load! NAME): writes/reads
+# /tmp/<name>.portal so the JS side can round-trip blobs to vault.
+import os
+os.makedirs('/tmp', exist_ok=True)
+
+def _portal_snapshot_prim(args, env):
+    if not args or not isinstance(args[0], str):
+        raise lumbda.LispErr('portal-snapshot!: expected string name')
+    path = f'/tmp/{args[0]}.portal'
+    g = env.g if env.g else _env
+    lumbda.portal_save(g, path)
+    return path
+
+def _portal_load_prim(args, env):
+    if not args or not isinstance(args[0], str):
+        raise lumbda.LispErr('portal-load!: expected string name')
+    path = f'/tmp/{args[0]}.portal'
+    restored_env, _cont = lumbda.portal_resume(path, _env)
+    # Copy bindings from the restored env into our live global env so
+    # subsequent evals see them. portal_resume gives us back a fresh
+    # env with the loaded bindings; we merge into _env in-place.
+    for sym, val in restored_env.b.items():
+        _env.b[sym] = val
+    return True
+
+_env.define(lumbda.S('portal-snapshot!'), _portal_snapshot_prim)
+_env.define(lumbda.S('portal-load!'), _portal_load_prim)
+
+# Streaming stdout wrapper — every write() also calls back into JS so
+# the worker can postMessage chunks to the UI during the eval. The
+# StringIO behind it still captures everything for the final return.
+from js import _lumbdaPyEmitChunk as _js_emit_chunk
+class _StreamingStdout(io.StringIO):
+    def write(self, s):
+        n = super().write(s)
+        try:
+            _js_emit_chunk(s)
+        except Exception:
+            pass
+        return n
+
 def _lumbda_eval(src):
-    buf = io.StringIO()
+    buf = _StreamingStdout()
     old = sys.stdout
     sys.stdout = buf
     last = None
@@ -81,15 +131,39 @@ def _lumbda_eval(src):
         if out and not out.endswith("\\n"):
             out += "\\n"
         out += rep
+        try:
+            _js_emit_chunk(("" if out.endswith(rep) else "\\n") + rep)
+        except Exception:
+            pass
     return out
 `);
 
     return {
-        async evalLisp(src) {
+        async evalLisp(src, onChunk) {
             pyodide.globals.set("_src_in", src);
-            return await pyodide.runPythonAsync("_lumbda_eval(_src_in)");
+            refs.currentOnChunk = onChunk || null;
+            try {
+                return await pyodide.runPythonAsync("_lumbda_eval(_src_in)");
+            } finally {
+                refs.currentOnChunk = null;
+            }
         },
         setBendUrl(url) { refs.bendUrl = url || null; },
+        // Portal save/resume bridge to vault — same shape as the C tier
+        // loader. Pyodide.FS is the same Emscripten FS, so the path
+        // /tmp/<name>.portal is reachable from JS exactly as in C.
+        portalSave(name) {
+            const path = `/tmp/${name}.portal`;
+            try {
+                const bytes = pyodide.FS.readFile(path);
+                return new TextDecoder().decode(bytes);
+            } catch (e) { return null; }
+        },
+        portalLoad(name, blob) {
+            const path = `/tmp/${name}.portal`;
+            try { pyodide.FS.mkdir("/tmp"); } catch (e) { /* exists */ }
+            pyodide.FS.writeFile(path, blob);
+        },
         heapStats() {
             // Pyodide's runtime memory is the Emscripten linear memory.
             // CPython's GC reclaims behind the scenes, so this number
diff --git a/wasm/python/lumbda-py.js b/wasm/python/lumbda-py.js
index 609cefc..34aff77 100644
--- a/wasm/python/lumbda-py.js
+++ b/wasm/python/lumbda-py.js
@@ -29,7 +29,15 @@ async function _bootstrap() {
     // the response text back. Sync XHR is the only sync HTTP available
     // in a Web Worker; perfect for the blocking eval model lumbda
     // primitives expect.
-    const refs = { bendUrl: null };
+    const refs = { bendUrl: null, currentOnChunk: null };
+    // Streaming output bridge — sys.stdout in pyodide is replaced
+    // (during _lumbda_eval) with a class whose write() calls back
+    // here. We forward to the current onChunk so the worker can
+    // postMessage chunks to the UI as work happens, instead of the
+    // user staring at a blank panel during a long bend dispatch.
+    globalThis._lumbdaPyEmitChunk = (s) => {
+        if (refs.currentOnChunk && s) refs.currentOnChunk(s);
+    };
     globalThis._lumbdaPyBendCall = (payload) => {
         if (!refs.bendUrl) return "no bend URL configured";
         try {
@@ -63,8 +71,50 @@ def _bend_call_prim(args, env):
     return str(_js_bend_call(payload))
 _env.define(lumbda.S('bend!-call'), _bend_call_prim)
 
+# Portal save/resume — REPL bridge. Mirrors the C tier's
+# (portal-snapshot! NAME) / (portal-load! NAME): writes/reads
+# /tmp/<name>.portal so the JS side can round-trip blobs to vault.
+import os
+os.makedirs('/tmp', exist_ok=True)
+
+def _portal_snapshot_prim(args, env):
+    if not args or not isinstance(args[0], str):
+        raise lumbda.LispErr('portal-snapshot!: expected string name')
+    path = f'/tmp/{args[0]}.portal'
+    g = env.g if env.g else _env
+    lumbda.portal_save(g, path)
+    return path
+
+def _portal_load_prim(args, env):
+    if not args or not isinstance(args[0], str):
+        raise lumbda.LispErr('portal-load!: expected string name')
+    path = f'/tmp/{args[0]}.portal'
+    restored_env, _cont = lumbda.portal_resume(path, _env)
+    # Copy bindings from the restored env into our live global env so
+    # subsequent evals see them. portal_resume gives us back a fresh
+    # env with the loaded bindings; we merge into _env in-place.
+    for sym, val in restored_env.b.items():
+        _env.b[sym] = val
+    return True
+
+_env.define(lumbda.S('portal-snapshot!'), _portal_snapshot_prim)
+_env.define(lumbda.S('portal-load!'), _portal_load_prim)
+
+# Streaming stdout wrapper — every write() also calls back into JS so
+# the worker can postMessage chunks to the UI during the eval. The
+# StringIO behind it still captures everything for the final return.
+from js import _lumbdaPyEmitChunk as _js_emit_chunk
+class _StreamingStdout(io.StringIO):
+    def write(self, s):
+        n = super().write(s)
+        try:
+            _js_emit_chunk(s)
+        except Exception:
+            pass
+        return n
+
 def _lumbda_eval(src):
-    buf = io.StringIO()
+    buf = _StreamingStdout()
     old = sys.stdout
     sys.stdout = buf
     last = None
@@ -81,15 +131,39 @@ def _lumbda_eval(src):
         if out and not out.endswith("\\n"):
             out += "\\n"
         out += rep
+        try:
+            _js_emit_chunk(("" if out.endswith(rep) else "\\n") + rep)
+        except Exception:
+            pass
     return out
 `);
 
     return {
-        async evalLisp(src) {
+        async evalLisp(src, onChunk) {
             pyodide.globals.set("_src_in", src);
-            return await pyodide.runPythonAsync("_lumbda_eval(_src_in)");
+            refs.currentOnChunk = onChunk || null;
+            try {
+                return await pyodide.runPythonAsync("_lumbda_eval(_src_in)");
+            } finally {
+                refs.currentOnChunk = null;
+            }
         },
         setBendUrl(url) { refs.bendUrl = url || null; },
+        // Portal save/resume bridge to vault — same shape as the C tier
+        // loader. Pyodide.FS is the same Emscripten FS, so the path
+        // /tmp/<name>.portal is reachable from JS exactly as in C.
+        portalSave(name) {
+            const path = `/tmp/${name}.portal`;
+            try {
+                const bytes = pyodide.FS.readFile(path);
+                return new TextDecoder().decode(bytes);
+            } catch (e) { return null; }
+        },
+        portalLoad(name, blob) {
+            const path = `/tmp/${name}.portal`;
+            try { pyodide.FS.mkdir("/tmp"); } catch (e) { /* exists */ }
+            pyodide.FS.writeFile(path, blob);
+        },
         heapStats() {
             // Pyodide's runtime memory is the Emscripten linear memory.
             // CPython's GC reclaims behind the scenes, so this number
diff --git a/www/playground/app.js b/www/playground/app.js
index 33736fc..2c61f8a 100644
--- a/www/playground/app.js
+++ b/www/playground/app.js
@@ -213,19 +213,28 @@ function runOnTierInWorker(tier, src, onLoading) {
     return new Promise((resolve, reject) => {
         const w = ensureWorker(tier);
         const myRunId = ++workerState.nextRunId;
+        let liveBlock = null;
         workerState.pending[tier] = { runId: myRunId, resolve, reject };
         const handler = (e) => {
             if (e.data.runId !== myRunId) return;
             if (e.data.kind === "loading") {
                 onLoading && onLoading(e.data.tier);
+            } else if (e.data.kind === "chunk") {
+                // First chunk spawns the in-progress block; subsequent
+                // chunks append to it. User sees displays land
+                // immediately, not just at the end of the run.
+                if (!liveBlock) liveBlock = startLiveBlock(tier);
+                liveBlock.append(e.data.chunk);
             } else if (e.data.kind === "done") {
                 w.removeEventListener("message", handler);
                 workerState.pending[tier] = null;
-                resolve(e.data.output);
+                resolve({ output: e.data.output, liveBlock });
             } else if (e.data.kind === "error") {
                 w.removeEventListener("message", handler);
                 workerState.pending[tier] = null;
-                reject(new Error(e.data.message));
+                const err = new Error(e.data.message);
+                err.liveBlock = liveBlock;
+                reject(err);
             }
         };
         w.addEventListener("message", handler);
@@ -233,6 +242,44 @@ function runOnTierInWorker(tier, src, onLoading) {
     });
 }
 
+// Spawn a tier-block in the output panel immediately on first streamed
+// chunk. Returns handles to append further chunks and to finalize the
+// timing header once the eval reports done. Mirrors appendBlock's
+// structure so finalized live blocks look identical to non-streamed
+// ones — same DOM, same CSS.
+function startLiveBlock(tierName) {
+    const block = document.createElement("div");
+    block.className = "tier-block";
+    const h = document.createElement("h3");
+    h.textContent = TIERS[tierName] || tierName;
+    const t = document.createElement("span");
+    t.className = "time";
+    t.textContent = "  (running…)";
+    h.appendChild(t);
+    block.appendChild(h);
+    const pre = document.createElement("pre");
+    pre.textContent = "";
+    block.appendChild(pre);
+    outputEl.appendChild(block);
+    return {
+        append(chunk) {
+            pre.textContent += chunk;
+            outputEl.scrollTop = outputEl.scrollHeight;
+        },
+        finalize(elapsedMs, kind, fullOutput) {
+            if (kind === "cancelled") t.textContent = `  (cancelled @ ${elapsedMs.toFixed(0)} ms)`;
+            else t.textContent = `  (${elapsedMs.toFixed(0)} ms)`;
+            if (kind === "error") pre.className = "err";
+            // If the streamed chunks miss anything (e.g. asm tier
+            // which doesn't stream yet), reconcile with the full
+            // output. No-op when streaming captured everything.
+            if (fullOutput && pre.textContent !== fullOutput) {
+                pre.textContent = fullOutput;
+            }
+        },
+    };
+}
+
 function cancelCurrentRun() {
     for (const t of Object.keys(workerState.workers)) {
         if (workerState.workers[t]) {
@@ -307,20 +354,26 @@ async function runAll() {
         return runOnTierInWorker(t, src, (loadingTier) => {
             setStatus(`loading ${loadingTier}…`, "busy");
         })
-            .then((output) => {
+            .then(({ output, liveBlock }) => {
                 const elapsed = performance.now() - startTimes[t];
                 delete startTimes[t];
-                appendBlock(t, elapsed, output, "ok");
+                // Streamed tiers already painted via liveBlock; just
+                // finalize the timing header. Non-streaming tiers
+                // (today: asm) get a fresh appendBlock at the end.
+                if (liveBlock) liveBlock.finalize(elapsed, "ok", output);
+                else appendBlock(t, elapsed, output, "ok");
                 return { tier: t, ok: true, elapsed };
             })
             .catch((e) => {
                 const elapsed = performance.now() - startTimes[t];
                 delete startTimes[t];
                 if (e.message === "cancelled") {
-                    appendBlock(t, elapsed, "(cancelled)", "cancelled");
+                    if (e.liveBlock) e.liveBlock.finalize(elapsed, "cancelled");
+                    else appendBlock(t, elapsed, "(cancelled)", "cancelled");
                     return { tier: t, cancelled: true, elapsed };
                 }
-                appendBlock(t, elapsed, e.message || String(e), "error");
+                if (e.liveBlock) e.liveBlock.finalize(elapsed, "error", e.message);
+                else appendBlock(t, elapsed, e.message || String(e), "error");
                 return { tier: t, error: e.message, elapsed };
             });
     });
diff --git a/www/playground/c/lumbda-c.loader.js b/www/playground/c/lumbda-c.loader.js
index 64e78c2..181d4fe 100644
--- a/www/playground/c/lumbda-c.loader.js
+++ b/www/playground/c/lumbda-c.loader.js
@@ -13,12 +13,25 @@ async function _bootstrap() {
     const wasmDir = new URL("./", import.meta.url).href;
     const { default: createLumbdaC } = await import(/* @vite-ignore */ factoryURL);
 
+    // Streaming output: every printf in the wasm fires Module.print
+    // synchronously. We buffer normally for the final return, but if
+    // the current eval registered an onChunk callback we also call
+    // through immediately so the worker can postMessage a chunk to
+    // the UI as each line emerges (vs the user staring at a blank
+    // panel for 18 seconds during the heavy bend call).
     let outBuf = [];
     let errBuf = [];
+    let currentOnChunk = null;
     const module = await createLumbdaC({
         locateFile: (p) => wasmDir + p,
-        print: (line) => outBuf.push(line),
-        printErr: (line) => errBuf.push(line),
+        print: (line) => {
+            outBuf.push(line);
+            if (currentOnChunk) currentOnChunk(line);
+        },
+        printErr: (line) => {
+            errBuf.push(line);
+            if (currentOnChunk) currentOnChunk(line);
+        },
     });
 
     const _init = module.cwrap("lumbda_wasm_init", null, []);
@@ -56,10 +69,12 @@ async function _bootstrap() {
             const path = `/tmp/${name}.portal`;
             module.FS.writeFile(path, blob);
         },
-        async evalLisp(src) {
+        async evalLisp(src, onChunk) {
             outBuf = [];
             errBuf = [];
+            currentOnChunk = onChunk || null;
             const errPtr = _eval(src);
+            currentOnChunk = null;
             let errMsg = "";
             if (errPtr) {
                 errMsg = module.UTF8ToString(errPtr);
diff --git a/www/playground/python/lumbda-py.js b/www/playground/python/lumbda-py.js
index 609cefc..34aff77 100644
--- a/www/playground/python/lumbda-py.js
+++ b/www/playground/python/lumbda-py.js
@@ -29,7 +29,15 @@ async function _bootstrap() {
     // the response text back. Sync XHR is the only sync HTTP available
     // in a Web Worker; perfect for the blocking eval model lumbda
     // primitives expect.
-    const refs = { bendUrl: null };
+    const refs = { bendUrl: null, currentOnChunk: null };
+    // Streaming output bridge — sys.stdout in pyodide is replaced
+    // (during _lumbda_eval) with a class whose write() calls back
+    // here. We forward to the current onChunk so the worker can
+    // postMessage chunks to the UI as work happens, instead of the
+    // user staring at a blank panel during a long bend dispatch.
+    globalThis._lumbdaPyEmitChunk = (s) => {
+        if (refs.currentOnChunk && s) refs.currentOnChunk(s);
+    };
     globalThis._lumbdaPyBendCall = (payload) => {
         if (!refs.bendUrl) return "no bend URL configured";
         try {
@@ -63,8 +71,50 @@ def _bend_call_prim(args, env):
     return str(_js_bend_call(payload))
 _env.define(lumbda.S('bend!-call'), _bend_call_prim)
 
+# Portal save/resume — REPL bridge. Mirrors the C tier's
+# (portal-snapshot! NAME) / (portal-load! NAME): writes/reads
+# /tmp/<name>.portal so the JS side can round-trip blobs to vault.
+import os
+os.makedirs('/tmp', exist_ok=True)
+
+def _portal_snapshot_prim(args, env):
+    if not args or not isinstance(args[0], str):
+        raise lumbda.LispErr('portal-snapshot!: expected string name')
+    path = f'/tmp/{args[0]}.portal'
+    g = env.g if env.g else _env
+    lumbda.portal_save(g, path)
+    return path
+
+def _portal_load_prim(args, env):
+    if not args or not isinstance(args[0], str):
+        raise lumbda.LispErr('portal-load!: expected string name')
+    path = f'/tmp/{args[0]}.portal'
+    restored_env, _cont = lumbda.portal_resume(path, _env)
+    # Copy bindings from the restored env into our live global env so
+    # subsequent evals see them. portal_resume gives us back a fresh
+    # env with the loaded bindings; we merge into _env in-place.
+    for sym, val in restored_env.b.items():
+        _env.b[sym] = val
+    return True
+
+_env.define(lumbda.S('portal-snapshot!'), _portal_snapshot_prim)
+_env.define(lumbda.S('portal-load!'), _portal_load_prim)
+
+# Streaming stdout wrapper — every write() also calls back into JS so
+# the worker can postMessage chunks to the UI during the eval. The
+# StringIO behind it still captures everything for the final return.
+from js import _lumbdaPyEmitChunk as _js_emit_chunk
+class _StreamingStdout(io.StringIO):
+    def write(self, s):
+        n = super().write(s)
+        try:
+            _js_emit_chunk(s)
+        except Exception:
+            pass
+        return n
+
 def _lumbda_eval(src):
-    buf = io.StringIO()
+    buf = _StreamingStdout()
     old = sys.stdout
     sys.stdout = buf
     last = None
@@ -81,15 +131,39 @@ def _lumbda_eval(src):
         if out and not out.endswith("\\n"):
             out += "\\n"
         out += rep
+        try:
+            _js_emit_chunk(("" if out.endswith(rep) else "\\n") + rep)
+        except Exception:
+            pass
     return out
 `);
 
     return {
-        async evalLisp(src) {
+        async evalLisp(src, onChunk) {
             pyodide.globals.set("_src_in", src);
-            return await pyodide.runPythonAsync("_lumbda_eval(_src_in)");
+            refs.currentOnChunk = onChunk || null;
+            try {
+                return await pyodide.runPythonAsync("_lumbda_eval(_src_in)");
+            } finally {
+                refs.currentOnChunk = null;
+            }
         },
         setBendUrl(url) { refs.bendUrl = url || null; },
+        // Portal save/resume bridge to vault — same shape as the C tier
+        // loader. Pyodide.FS is the same Emscripten FS, so the path
+        // /tmp/<name>.portal is reachable from JS exactly as in C.
+        portalSave(name) {
+            const path = `/tmp/${name}.portal`;
+            try {
+                const bytes = pyodide.FS.readFile(path);
+                return new TextDecoder().decode(bytes);
+            } catch (e) { return null; }
+        },
+        portalLoad(name, blob) {
+            const path = `/tmp/${name}.portal`;
+            try { pyodide.FS.mkdir("/tmp"); } catch (e) { /* exists */ }
+            pyodide.FS.writeFile(path, blob);
+        },
         heapStats() {
             // Pyodide's runtime memory is the Emscripten linear memory.
             // CPython's GC reclaims behind the scenes, so this number
diff --git a/www/playground/runner.js b/www/playground/runner.js
index aa602af..5379f6f 100644
--- a/www/playground/runner.js
+++ b/www/playground/runner.js
@@ -29,9 +29,9 @@ export async function getTier(name, onLoad) {
     return cache[name];
 }
 
-export async function evalOnTier(name, src, onLoad) {
+export async function evalOnTier(name, src, onLoad, onChunk) {
     const tier = await getTier(name, onLoad);
-    return tier.evalLisp(src);
+    return tier.evalLisp(src, onChunk);
 }
 
 export function heapStats(name) {
@@ -39,3 +39,20 @@ export function heapStats(name) {
     if (!tier || !tier.heapStats) return null;
     return tier.heapStats();
 }
+
+// Portal — REPL save/resume bridge. portalSave reads a snapshot blob
+// from the tier (MEMFS on the C tier, similar bridges on others as
+// they land). Returns null when the tier hasn't implemented portals
+// yet (asm-wat today) so the caller can surface a friendly message
+// instead of crashing. portalLoad is the symmetric write-in path.
+export function portalSave(name, checkpointName) {
+    const tier = cache[name];
+    if (!tier || !tier.portalSave) return null;
+    return tier.portalSave(checkpointName);
+}
+export function portalLoad(name, checkpointName, blob) {
+    const tier = cache[name];
+    if (!tier || !tier.portalLoad) return false;
+    tier.portalLoad(checkpointName, blob);
+    return true;
+}
diff --git a/www/playground/worker.mjs b/www/playground/worker.mjs
index 760c680..795f0eb 100644
--- a/www/playground/worker.mjs
+++ b/www/playground/worker.mjs
@@ -3,7 +3,7 @@
 // live ms counter actually ticks AND so cancel works (main thread
 // terminates this worker via worker.terminate()).
 
-import { evalOnTier, setBendUrl, heapStats } from "./runner.js";
+import { evalOnTier, setBendUrl, heapStats, portalSave, portalLoad } from "./runner.js";
 
 self.onmessage = async (e) => {
     const { kind } = e.data;
@@ -17,12 +17,39 @@ self.onmessage = async (e) => {
         self.postMessage({ kind: "heap", runId: e.data.runId, tier: e.data.tier, stats: heapStats(e.data.tier) });
         return;
     }
+    if (kind === "portal-save") {
+        // Pulls the most-recent /tmp/<name>.portal out of the tier's
+        // MEMFS after the lisp side ran (portal-snapshot! NAME). Main
+        // thread encrypts and stuffs it into the vault.
+        const blob = portalSave(e.data.tier, e.data.name);
+        self.postMessage({ kind: "portal-save", runId: e.data.runId, name: e.data.name, blob });
+        return;
+    }
+    if (kind === "portal-load") {
+        // Hydrates MEMFS from a vault-decrypted blob so a subsequent
+        // (portal-load! NAME) eval finds the file ready.
+        const ok = portalLoad(e.data.tier, e.data.name, e.data.blob);
+        self.postMessage({ kind: "portal-load", runId: e.data.runId, name: e.data.name, ok });
+        return;
+    }
     if (kind !== "eval") return;
     const { runId, tier, src } = e.data;
     try {
-        const output = await evalOnTier(tier, src, (loadingTier) => {
-            self.postMessage({ kind: "loading", runId, tier: loadingTier });
-        });
+        const output = await evalOnTier(
+            tier,
+            src,
+            (loadingTier) => {
+                self.postMessage({ kind: "loading", runId, tier: loadingTier });
+            },
+            // Stream every print/display from the tier to the main
+            // thread as it happens. Sync XHR inside bend!-call still
+            // blocks the worker, but displays BEFORE/AFTER the bend
+            // round-trip surface immediately instead of waiting for
+            // the whole eval to finish. Long demos feel alive.
+            (chunk) => {
+                self.postMessage({ kind: "chunk", runId, tier, chunk });
+            },
+        );
         self.postMessage({ kind: "done", runId, output });
     } catch (err) {
         self.postMessage({ kind: "error", runId, message: err && err.message ? err.message : String(err) });