Benchmarks

Benchmark score in milliseconds per operation — lower is better. Every chart is the checked-in baseline from the repo’s own suites: same apps, same hardware, same run. Every primary React comparison uses the official React Compiler; memo-wall also shows an explicitly uncompiled control. Reproduce them with pnpm bench:all, and see benchmarks/ for what each suite measures.

Every suite at a glance

Start here: each cell is the geometric mean of that suite’s per-operation scores relative to Octane — 1× is Octane, below 1× is faster than Octane. Pick frameworks and a suite for the bar view, or scan the whole grid and re-baseline it to each suite’s fastest framework. The sections below have the full per-operation charts in milliseconds.

frameworks

suite

1× Octane

Octane (.tsrx)

1×

Vue Vapor 3.6 beta

1.2×

Solid 2.0 beta

1.7×

Ripple 0.3

1.7×

Svelte 5

2.7×

Preact 10

3.6×

React 19 + Compiler

4.1×

all suites

suite	Octane (.tsrx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
js-framework	1×	4.1×	3.6×	1.7×	2.7×	1.7×	1.2×
todomvc	1×	3.3×	3.3×	1.8×	1.6×	1.1×	1.2×
weather-app	1×	1.3×	1.5×	1.3×	1.1×	—	—
weather-app-lighthouse	1×	1.1×	0.96×	0.94×	0.96×	—	—
chat-stream	1×	3.6×	2.8×	1.5×	2.1×	1.2×	1.1×
svg-dashboard	1×	1.2×	—	1.4×	1.1×	—	—
js-framework-reorder	1×	2.9×	6.0×	1.5×	2.2×	1.5×	2.2×
dbmon	1×	1.8×	1.9×	2.8×	1.2×	1.2×	1.1×
effectful-list	1×	0.72×	3.6×	0.59×	0.74×	1.0×	0.82×
memo-wall	1×	3.7×	16×	0.95×	2.9×	5.9×	0.85×
recursive-context	1×	1.5×	1.3×	1.0×	2.1×	1.2×	1.1×
spa-navigation	1×	0.79×	—	0.86×	—	—	1.2×
signal-favoring	1×	3.0×	3.0×	0.55×	0.77×	0.21×	0.30×
portal-swarm	1×	2.2×	6.1×	1.0×	2.5×	2.7×	1.1×
async-waterfall	1×	11×	8.7×	0.91×	0.92×	0.89×	—
news	1×	3.0×	2.0×	1.9×	1.0×	1.7×	1.3×
streaming-ssr	1×	0.81×	1.2×	3.1×	—	0.68×	—
bundle-size	1×	3.1×	0.41×	0.71×	0.96×	0.73×	1.3×
ssr-throughput	1×	2.4×	2.3×	1.6×	0.98×	1.4×	0.86×
geomean	1×	2.1×	2.6×	1.2×	1.4×	1.2×	1.0×

Each cell is that framework’s score relative to Octane (1×), geometric mean across the suite’s operations. Green is faster, red is slower.

FasterSlowerOctane = 1× baseline

Octane vs the field

Octane compiled from .tsrx and .tsx, against React 19 with the official React Compiler, Preact 10, Solid, Svelte 5, Ripple and Vue Vapor — each framework’s fixture written the way its docs recommend. React Compiler is enabled for every primary React comparison; memo-wall also includes a clearly labeled uncompiled React control. The ranked chips on each card are that suite’s one-line reading: every framework’s geometric mean across the card’s operations, relative to Octane, fastest first.

js-frameworkkrausest-style table operations over 1,000 rows — create, replace, partial update, select, swap, remove, clear.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 8 iterations)

operation	Octane (.tsrx)	Octane (.tsx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
run	2.4	2.9	5.8	7.6	2.1	3.8	2.4	2.7
replace	5.1	5.3	10	13	5.2	7.1	5.0	5.2
add	2.0	2.1	4.5	6.5	1.7	3.2	1.9	2.0
update	0.34	0.46	0.82	1.3	0.98	0.66	0.32	0.42
select	0.18	0.38	0.62	0.82	0.54	1.1	0.62	0.24
swap	0.42	0.74	4.5	0.66	0.52	0.60	0.50	0.50
remove	0.22	0.26	0.62	0.50	0.16	0.48	0.22	0.22
runlots	15	16	115	52	14	24	14	16
select_lots	0.06	0.88	2.1	2.4	2.2	4.5	3.6	0.12
clear	21	21	39	31	22	25	23	21

todomvcTodoMVC workflows — add, complete, filter, edit, clear and destroy items, with native form/input events in the loop.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 8 iterations)

operation	Octane (.tsrx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
add 100	1.5	13	11	2.3	3.3	2.3	2.0
toggle all on	0.34	0.40	0.46	1.1	0.44	0.34	0.32
toggle all off	0.26	0.44	0.46	1.0	0.42	0.34	0.36
complete 25	0.82	5.9	5.8	1.1	1.2	0.94	0.78
filter cycle	0.86	1.2	1.3	1.00	1.1	0.64	1.1
edit 10	1.9	5.3	5.3	1.8	2.9	1.2	1.5
clear completed	0.24	0.42	0.48	0.34	0.28	0.20	0.38
destroy 25	0.36	5.7	5.1	0.92	1.2	0.72	0.78

weather-appA real weather dashboard — cold start to interactive, keyed forecast churn, and async city search with error and recovery flows. The DOM-census counters in the baseline are diagnostics and stay out of the chart.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 8 iterations)

operation	Octane (.tsrx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Vue 3.6
cold ready	24	33	22	23	24	26
forecast churn	3.2	5.1	8.8	7.9	3.9	4.3
search city	1.5	1.6	2.0	2.4	1.7	1.8
search error	0.43	0.79	0.72	0.36	0.40	0.48
search recover	1.2	1.3	1.3	1.3	1.2	1.4

weather-app — LighthouseDesktop Lighthouse lab metrics for the same weather app — simulated and observed first and largest contentful paint, speed index, and total blocking time. Simulated paints use Lighthouse’s desktop-network model; observed paints come from the unthrottled browser trace. Cumulative layout shift is unitless and omitted from the millisecond chart.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 5 iterations)

operation	Octane (.tsrx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Vue 3.6
simulated FCP	323	363	285	283	284	323
observed FCP	66	85	74	81	77	78
simulated LCP	365	404	326	329	370	370
observed LCP	93	102	95	81	84	94
speed index	323	363	285	283	284	323
TBT	0	0	0	0	0	0

chat-streamChat UI workloads — token streaming, coarse updates, history append, conversation switches and text input.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 8 iterations)

operation	Octane (.tsrx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
fine stream	1.3	6.8	4.3	2.8	3.0	1.2	1.2
coarse stream	0.68	2.2	1.5	1.1	1.3	0.76	0.68
append history	0.72	7.8	5.1	1.4	2.6	1.6	1.1
switch conversation	4.2	5.0	4.0	4.9	9.7	4.2	4.1
type 160 chars	1.1	3.0	4.2	1.1	1.5	1.2	1.2

svg-dashboardA hand-rolled SVG observability dashboard — path and transform churn, keyed reconciliation inside SVG, foreignObject labels, portal tooltips, runtime icons, and style/spread updates.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 20 iterations)

operation	Octane (.tsrx)	React 19 + Compiler	Solid 2.0 beta	Svelte 5
mount	2.8	3.7	5.9	3.8
chart tick	2.8	4.0	3.7	3.4
sparse tick	0.79	2.0	2.1	2.1
drag nodes	6.7	3.6	5.6	3.3
pan and zoom	1.7	4.9	1.0	2.4
toggle selection	1.8	1.1	3.3	1.8
topology churn	1.8	1.6	3.8	2.0
label churn	1.6	1.7	1.3	1.3
tooltip swarm	1.4	3.1	1.5	1.9
icon swap	2.1	2.4	3.6	2.7
series toggle	2.2	2.7	3.0	2.5
style/spread pulse	2.7	1.3	5.8	2.0

js-framework-reorderThe keyed-reorder matrix — reverse, shuffle, rotations, prepends and displacements — stressing the keyed reconciler. Ripple’s prepend and insert-mid cells are omitted because their current identity gates fail; invalid timings are never charted.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 8 iterations)

operation	Octane (.tsrx)	Octane (.tsx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
reverse	0.95	1.1	2.0	2.4	1.0	14	0.96	1.2
shuffle	1.1	1.3	1.7	2.2	1.2	1.6	1.2	1.4
rotatef	0.05	0.13	1.3	0.12	0.08	0.11	0.09	0.58
rotateb	0.04	0.12	0.11	0.12	0.07	0.08	0.08	0.13
prepend100	0.36	0.38	0.90	1.1	0.22	0.58	—	0.32
append100	0.32	0.38	0.64	1.0	0.26	0.58	0.36	0.36
insertmid100	0.28	0.42	0.78	1.1	0.28	0.60	—	0.36
removefirst	0.02	0.08	0.13	0.17	0.06	0.10	0.06	0.03
removeevery10	0.09	0.14	0.25	0.25	0.16	0.14	0.10	0.21
displace3	0.06	0.14	0.14	1.3	0.13	0.12	0.10	0.19
displace4	0.07	0.14	0.14	1.3	0.13	0.11	0.11	0.20
displace5	0.08	0.14	0.15	1.3	0.14	0.12	0.11	0.20
displace6	0.06	0.14	0.15	1.3	0.14	0.12	0.11	0.20
displace8	0.08	0.15	0.16	1.2	0.14	0.12	0.12	0.20

dbmonThe DBMonster dashboard — high-frequency cell updates across a wall of database rows.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 30 iterations)

operation	Octane (.tsrx)	Octane (.tsx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
mount	4.1	5.0	6.7	6.5	15	5.2	6.1	5.0
tick	1.4	1.6	2.2	3.2	9.3	2.0	2.0	1.6
tick_partial	0.45	0.55	1.2	2.0	1.4	0.61	0.59	0.53
remount	4.1	4.4	7.3	6.9	12	4.5	4.4	4.0
sort	2.1	2.3	4.0	4.4	3.0	2.5	2.1	2.2
unmount	1.8	1.6	2.3	1.7	2.8	1.7	1.6	1.7

effectful-listA 1,000-item list where every row runs effects and refs — the subsystems a plain row bench never touches.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 30 iterations)

operation	Octane (.tsrx)	Octane (.tsx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
mount_1k	8.3	8.9	10	12	9.9	9.4	9.3	11
update_nodeps	0.14	0.26	0.00	7.2	0.00	0.01	0.08	0.00
update_deps	2.1	2.3	2.7	7.5	2.7	2.3	2.4	2.6
clear	0.91	0.94	1.4	0.97	1.5	1.1	1.2	1.8
remount	8.8	9.2	11	14	11	10	11	13
remove_100_scattered	0.36	0.52	0.70	1.9	0.84	0.40	0.44	1.0

memo-wallMemo bail-out walls — parent re-renders against memoized subtrees, and context updates punching through them. The primary React entry uses React Compiler; an additional, explicitly uncompiled React control isolates the compiler’s effect. Solid, Svelte, Ripple and Vue Vapor have no parent re-render to absorb, so their near-zero wall ops are the fine-grained model’s honest number.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 20 iterations)

operation	Octane (.tsrx)	Octane (.tsx)	React 19 + Compiler	React 19 (uncompiled control)	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
mount	11	15	12	11	15	11	11	10	14
parent_rerender_equal_A	0.00	0.00	0.00	0.15	0.16	0.00	0.01	0.05	0.00
parent_rerender_equal_B	0.00	0.18	0.00	0.18	0.19	0.00	0.01	0.05	0.00
one_change_A	0.00	0.04	0.16	0.17	0.17	0.03	0.05	0.07	0.02
one_change_B	0.16	0.25	0.21	0.20	0.20	0.02	0.05	0.07	0.02
ctx_through_wall_A	0.32	0.57	0.77	0.91	1.0	0.29	0.36	0.34	0.22
ctx_through_wall_B	0.33	0.71	0.76	0.91	1.0	0.29	0.37	0.33	0.22

recursive-contextA deep recursive tree driven by context updates — mount, root and partial updates, unmount.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 20 iterations)

operation	Octane (.tsrx)	Octane (.tsx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
mount	5.0	11	5.3	3.6	4.6	14	5.1	10
update_root	0.73	3.3	0.68	1.1	0.55	0.45	0.56	0.37
update_partial	0.03	0.16	0.07	0.11	0.04	0.04	0.05	0.01
partial_unmount	0.02	0.04	0.05	0.01	0.01	0.09	0.03	0
partial_remount	0.07	0.20	0.09	0.14	0.16	0.38	0.18	0.15
unmount	0.46	2.2	0.73	0.48	0.46	0.76	0.34	0.68

spa-navigationFull-page client navigation — routed-subtree teardown and mount while the app shell survives, including nested-layout reuse and a 6× CPU-throttled route swap.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 20 iterations)

operation	Octane (.tsrx)	Octane (.tsx)	React 19 + Compiler	Solid 2.0 beta	Vue Vapor 3.6 beta
deep route swap	2.4	7.7	1.5	1.8	2.6
deep → nested	1.2	3.2	0.81	0.54	0.75
nested → deep	1.5	4.1	1.3	1.8	2.6
nested route swap	0.19	0.47	0.23	0.24	0.31
deep route swap (6× CPU)	15	46	9.9	13	18

signal-favoringDeep-tree state bumps at increasing depths — the workload shape signal frameworks are built around.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 20 iterations)

operation	Octane (.tsrx)	Octane (.tsx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
mount	2.3	2.4	2.9	1.2	1.7	1.7	1.8	1.7
bump_shallow	0.01	0.01	0.00	0.05	0.00	0.00	0.00	0.00
bump_middle	0.00	0.00	0.02	0.03	0.00	0.00	0.00	0.00
bump_deep	0.00	0.00	0.04	0.01	0.00	0.00	0.00	0.00
bump_sweep	0.03	0.04	0.16	0.23	0.02	0.02	0.01	0.01
bump_sweep_batched	0.02	0.02	0.06	0.05	0.01	0.01	0.00	0.00
bump_sweep_reverse	0.02	0.02	0.06	0.06	0.01	0.01	0.00	0.00
unmount	0.15	0.13	0.20	0.13	0.13	0.15	0.12	0.15

portal-swarmMany portals mounting, opening, closing and re-rendering — dispatching through portal boundaries.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 20 iterations)

operation	Octane (.tsrx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
mount_closed	3.3	4.0	3.0	4.2	4.2	2.3	4.1
open_all	1.4	1.4	3.3	1.2	2.8	3.4	2.1
rerender_open_A	0	0.01	0.49	0	0.01	0.03	0.00
rerender_open_B	0.00	0.01	0.49	0.00	0.02	0.03	0.00
rerender_open_B_stable	0.00	0.01	0.26	0.01	0.02	0.03	0
open_close_cycle	1.7	1.8	3.4	1.3	2.7	2.3	2.2
open_close_distinct	1.9	1.9	3.5	1.4	2.8	2.5	2.3
dispatch_through_portal	0.00	0.01	0.00	0.00	0.00	0.00	0.00

async-waterfallTen nested async levels with 16ms simulated latency — Octane’s compiled parallel-use path versus React and Preact nested-use waterfalls and signal-first models.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 10 iterations)

operation	Octane (.tsrx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3
init	21	305	180	19	19	19
update	20	184	180	19	19	18

newsA news-site page: full SSR render and client hydration of the same app.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 20 iterations)

operation	Octane (.tsrx)	Octane (.tsx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
SSR render	0.02	0.03	0.08	0.07	0.05	0.02	0.05	0.02
hydrate	1.4	1.9	2.5	1.3	1.8	1.4	1.2	1.6

streaming-ssrStreaming SSR shell and completion times for staggered Suspense and all-fast renders. Preact participates with its public stream renderer; Svelte 5 is omitted because its public server renderer is buffered.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 30 iterations)

operation	Octane (.tsrx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Ripple 0.3
staggered shell	0.31	0.44	0.58	2.4	0.35
staggered complete	51	51	51	51	51
all-fast shell	0.17	0.06	0.21	1.9	0.05
all-fast complete	1.8	1.5	1.5	2.0	1.1

bundle-sizeProduction shipped JavaScript bytes with normalized minification — total gzip across the rows, TodoMVC, chat and weather fixtures.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (production build bytes)

operation	Octane (.tsrx)	Octane (.tsx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta	Vue 3.6
rows total gzip	18 kB	35 kB	62 kB	10 kB	14 kB	18 kB	13 kB	—	—
TodoMVC total gzip	18 kB	—	60 kB	7 kB	13 kB	18 kB	13 kB	24 kB	—
chat total gzip	19 kB	—	61 kB	7 kB	14 kB	19 kB	13 kB	25 kB	—
weather total gzip	28 kB	—	67 kB	10 kB	17 kB	24 kB	—	—	31 kB

ssr-throughputServer renders of the news page at 50 and 500 items — benchmark score ms per render.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 10 iterations)

operation	Octane (.tsrx)	React 19 + Compiler	Preact 10	Solid 2.0 beta	Svelte 5	Ripple 0.3	Vue Vapor 3.6 beta
news-50	0.04	0.10	0.09	0.07	0.04	0.05	0.04
news-500	0.45	1.1	1.0	0.67	0.43	0.66	0.38

Head-to-head targets

Suites whose roster is a specific stack rather than the whole field: Octane against React on real HTTP servers, composed async trees and the same TanStack Start app, and the Octane Three renderer against React Three Fiber and hand-written Three.js. These cards stay out of the at-a-glance grid above, which aggregates only the full-roster suites.

async-compositionComposed async use() trees — Octane’s compiled parallel-use batching against React’s render-and-suspend loop. The wave, call and span counts in the baseline are deterministic diagnostics and stay out of the chart.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 10 iterations)

operation	Octane (.tsrx)	React 19 + Compiler
initial render	109	318
update	107	160

ssr-httpReal production HTTP servers rendering the news page — renderer import, cold spawn/listen/first-byte, and streamed shell and completion times over HTTP for staggered and all-fast Suspense.

geometric mean across all operations, × vs Octane (.tsrx) — lower is better

Data table (score ms, 10 iterations)

operation	Octane (.tsrx)	React 19 + Compiler
import renderer	1.3	8.6
cold spawn→listen	32	39
cold listen→first byte	4.5	5.1
cold spawn→first byte	37	44
staggered shell	1.1	1.3
staggered complete	52	52
all-fast shell	0.28	0.25
all-fast complete	1.7	1.7

tanstack-startThe same TanStack Start app served by React and by the Octane port — cold start to first byte, warm TTFB and completion for the posts and deferred routes, the deferred stream tail, and sequential home requests.

geometric mean across all operations, × vs Octane Start (minimal server) — lower is better

Data table (score ms, 7 iterations)

operation	Octane Start (minimal server)	Octane Start (nitro)	React (TanStack Start)
cold spawn→listen	40	50	61
cold listen→first byte	15	23	15
cold spawn→first byte	55	73	76
warm TTFB /posts	1.3	1.4	0.78
warm total /posts	1.3	1.4	0.80
warm TTFB /deferred	2.9	2.9	2.8
warm total /deferred	86	85	84
deferred stream tail	83	82	81
sequential home requests	0.87	0.85	0.62

three-rendererObject lifecycle at 1,000 meshes — mount, update, reorder, unmount, direct and component-driven reconstruction with disposal, a frame with 1,000 subscribers, and raycast events — for the Octane Three renderer, React Three Fiber, and hand-written Three.js.

geometric mean across all operations, × vs Octane Three — lower is better

Data table (score ms, 20 iterations)

operation	Octane Three	React Three Fiber 9	Plain Three.js
mount 1k	2.2	2.7	1.2
update 1k	0.43	0.46	0.01
reorder 1k	1.2	1.9	0.46
unmount 1k	0.79	2.7	0.31
reconstruct + dispose 1k	1.3	1.6	1.1
reconstruct component + dispose 1k	1.4	1.5	1.1
frame, 1k subscribers	0.03	0.03	0.03
raycast event	0.16	0.32	0.08

three-bundle-sizeShipped gzip JavaScript for a minimal scene and a helpers-heavy scene across the three renderers.

geometric mean across all operations, × vs Octane Three — lower is better

Data table (production build bytes)

operation	Octane Three	React Three Fiber 9	Plain Three.js
minimal scene	272 kB	184 kB	38 kB
full helpers	272 kB	244 kB	186 kB

Octane on Lynx

The dual-thread Lynx renderer on the same cross-framework table app as ReactLynx and both Vue Lynx renderers, measured two ways: wall clock on the Lynx-for-Web host with one shared page driver, and the deterministic wire cost of each commit — command counts against the changed-rows floor, exact and machine-independent, gated in CI. The gap between the wire bars is a KNOWN GAP the lynx-table ratio guards track; fixes tighten the guards and close the bars.

lynx-table — Lynx for Web wall clockThe cross-framework table on the Lynx-for-Web host at 10,000 rows — Octane, ReactLynx, and both Vue Lynx renderers driven by the byte-identical page driver, medians over fresh-page reps. Host-bound milliseconds; the ratios are the portable claim.

geometric mean across all operations, × vs Octane on Lynx — lower is better

Data table (score ms, 3 iterations)

operation	Octane on Lynx	ReactLynx 0.122	Vue Lynx vdom 3.6 beta	Vue Lynx vapor 3.6 beta
create 10k rows	4368	1826	2245	2342
update every 10th	925	270	196	166
select one row	752	160	149	148
update storm ×50	3288	6259	3324	1501
select storm ×30	941	2579	1642	147

lynx-table — commit wire costHost commands per operation crossing the background→main thread wire for the 10,000-row cross-framework table, against the changed-rows floor: the commands a change of that size strictly implies. The gap between the two bars is the tracked, CI-guarded optimization target.

geometric mean across all operations, × vs Octane on Lynx — lower is better

Data table (deterministic counts)

operation	Octane on Lynx	Changed-rows floor
create 10k rows	160,000	160,000
update every 10th	1,000	1,000
select one row	2	2
update storm ×50	50,000	50,000
select storm ×30	60	60

The authoring cliff

Octane against itself: what naively-authored code costs relative to the compiled fast path. The de-opt fixtures render the same apps through plain createElement — the shape every binding produces — so the cliff is a first-class, regression-guarded number.

js-framework — the authoring cliffThe same 1,000-row app authored four ways: tuned .tsrx, React-style naive .tsrx and .tsx, and plain-.ts createElement with no compiler involvement (the shape every binding produces).

geometric mean across all operations, × vs Tuned .tsrx — lower is better

Data table (score ms, 8 iterations)

operation	Tuned .tsrx	Naive .tsrx (React-style)	Naive .tsx	Plain .ts createElement
run	2.6	3.9	4.3	5.7
replace	4.9	5.9	6.3	9.5
add	1.9	3.1	3.6	8.7
update	0.38	1.3	1.5	6.2
select	0.20	1.5	1.6	7.0
swap	0.46	1.5	1.7	8.3
remove	0.22	0.90	1.1	4.2
runlots	15	22	24	37
select_lots	0.12	6.3	8.5	39
clear	21	21	23	36

dbmon — de-opt pathThe exact dbmon workload in plain-.ts createElement (full descriptor reconciliation) against the compiled fixture.

geometric mean across all operations, × vs Compiled .tsrx — lower is better

Data table (score ms, 30 iterations)

operation	Compiled .tsrx	Plain .ts createElement
mount	4.0	6.8
tick	1.4	5.1
tick_partial	0.44	4.0
remount	4.0	6.5
sort	2.0	6.4
unmount	1.6	1.8

SSR scenariosOctane-only SSR shapes — suspense waterfalls by depth, a 32-wide fan-out, an escape-heavy page, and the de-opt page pair. Benchmark score ms per render.

Data table (score ms, 10 iterations)

operation	Octane SSR
waterfall depth 1	0.04
waterfall depth 2	0.05
waterfall depth 4	0.07
depth 4 × 32 wide	2.3
escape-heavy page	3.2
de-opt page (fast)	0.96
de-opt page (de-opted)	2.3