Performance & Benchmarks

Open benchmarks.
All runs linked.

gRPC streaming latency. Transaction landing times. Full distribution — P50, P95, P99. Every test has a date, a node, a run count, and a public link.

Start for Free Read the Docs Aperture TxStream · Beam · gRPC benchmarks

82%

TxStream first vs Jito ShredStreammedian lead +284 µs · 44,574 tx

68.4%

TxStream first vs Yellowstonemedian lead +62 µs · full decoded tx

330ms

Beam avg landingvs 457ms standard RPC

100%

Beam success rateall providers · 10 runs each

Aperture TxStream · gRPC Streaming

First in 82% of races
vs Jito ShredStream.

TxStream reconstructs, decodes and streams transactions straight from shreds — before validator replay. On July 17 we raced it against Jito ShredStream and Yellowstone gRPC on 44,574 shared transactions with 100% coverage: every transaction was seen by every feed.

TxStream vs Jito ShredStream · 44,574 shared tx

Aperture TxStream — winner

Races won

82%

of 44,574 shared transactions

Median lead

+284 µs

Payload

Fully decoded tx

server-side filters applied

Jito ShredStream — raw shreds

Races won

18%

Median deficit

284 µs

Payload

Raw shreds

you deshred + decode yourself

TxStream vs Yellowstone gRPC · same run

Aperture TxStream — winner

Races won · full decoded tx

68.4%

median lead +62 µs

Races won · signatures-only mode

99.97%

median lead +7.77 ms

Yellowstone — standard gRPC

Races won · full decoded tx

31.6%

Races won · vs signatures-only

0.03%

Head-to-head against every endpoint in the run, TxStream was the fastest feed in 65.1% of races — with a median lag of 0 µs when it wasn't first.

Race win rate per matchup

% of races TxStream delivered the transaction first · 44,574 shared tx · July 17, 2026

Median lead when TxStream won

Full decoded payload mode · signatures-only mode adds +7.77 ms median lead vs Yellowstone

Same race — different payload

TxStream wins these races while doing more work per transaction: it deshreds, decodes, filters server-side — and runs real-time execution simulation — before the data leaves our edge. ShredStream hands you raw shreds; the deshredding and decoding time moves onto your stack.

Jito ShredStream · raw firehose

Races won 18%

Payload Raw shreds

Server-side filters No

Deshred + decode On your stack

Tx simulation No

Aperture TxStream · decoded feed

Races won 82%

Payload Fully decoded tx

Server-side filters Yes

Deshred + decode Done server-side

Tx simulation Real-time, ~95% accuracy

Real-time simulation, in-stream: TxStream can attach a predicted execution result to each transaction — predicted to land or to fail — at ~95% agreement with actual execution, adding 791 µs median to delivery. Included in the Aperture plan.

WebSocket · Independent Benchmark

First to deliver the slot.
Measured from the outside.

CompareNodes runs RPC Inspector Pro — independent probes that benchmark public RPC endpoints from 30 global locations.

RPC Fast WSS vs Alchemy WSS · slotSubscribe · 3 runs · 257 slots

RPC Fast WSS — Winner

First slot delivery

66%

170 / 257 slot observations

First or tied

84%

Win rate per run

60% · 64% · 74%

Success rate

100%

0 slot gaps · 0 rate limits

Alchemy — Standard WSS

First slot delivery

16%

42 / 257 slot observations

First or tied

34%

Win rate per run

18% · 18% · 13%

Success rate

100%

0 slot gaps · 0 rate limits

Per-slot margins are single-digit milliseconds — when Alchemy did take a slot first, the gap averaged 6–11ms. The signal is frequency: across all three runs, RPC Fast saw the slot first 4× more often.

Who delivered the slot first

Share of 257 slot observations · slotSubscribe · Frankfurt · July 14, 2026

Full runs on CompareNodes: Run 1 ↗ Run 2 ↗ Run 3 ↗ Re-run it live ↗

Beam · Transaction Landing

Routes to the fastest
provider on every run.

Beam scores providers in real time and routes each transaction to the top performer. P90 is weighted 4× in the scoring — tail latency is where landing fails under load. Available on all paid plans (extra fees via partners' network).

Provider scoring · P90-weighted 4× · 10 runs each

beam-falcon avg 229ms · P90 444ms · ack 54ms

100.00%

beam-astralane avg 231ms · P90 444ms · ack 52ms

99.78%

beam-bloxroute avg 233ms · P90 444ms · ack 53ms

99.66%

All three providers: 100% success ratio · avg landed slots 0.7 · P90 landed slots 1.0

Beam vs standard RPC · 20 runs

Provider	Avg landed	P90 landed	Performance score	Avg ack
beam-astralane	330ms	589ms	97.42%	53ms
rpcfast-rpc (standard)	457ms	1,065ms	82.20%	28ms

Beam: −127ms avg · P90 gap −476ms. Standard RPC acks faster (no routing overhead) but landing latency is 38% higher.

Beam vs standard RPC — landing latency

Avg and P90 landing time · 20 runs

Test Conditions

Public methodology.
No exceptions.

Every number includes the test date, node spec, run count, and scoring formula. Runs are linked — check them yourself.

Test Environment

gRPC race date July 17, 2026
gRPC sample 44,574 shared tx · 100% coverage
gRPC metrics per-pair races · medians + p95s
WSS / Beam date May 19, 2026 — Frankfurt
WSS / Beam node Latitude.sh — f4.metal.small
Beam scoring 10 runs · vs RPC — 20 runs

Beam Scoring Formula

landed_ms 0.2 × avg + 0.8 × P90 (P90 weighted 4×)
landed_slots 0.2 × avg + 0.8 × P90
landed_idx 0.5 × avg + 0.5 × P90
same_slot higher is better (0–100)
success_ratio higher is better (0–100)
Final score unweighted average of all five

Open benchmarks.All runs linked.

First in 82% of racesvs Jito ShredStream.

Race win rate per matchup

Median lead when TxStream won

First to deliver the slot.Measured from the outside.

Who delivered the slot first

Routes to the fastestprovider on every run.

Beam vs standard RPC — landing latency

Public methodology.No exceptions.

Test Environment

Beam Scoring Formula

Test it yourself.

Open benchmarks.
All runs linked.

First in 82% of races
vs Jito ShredStream.

First to deliver the slot.
Measured from the outside.

Routes to the fastest
provider on every run.

Public methodology.
No exceptions.