> For the complete documentation index, see [llms.txt](https://docs.rpcfast.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.rpcfast.com/rpc-fast-saas-solana/performance-and-benchmarking.md).
# Performance and Benchmarking
## gRPC Performance
The latest [gRPC stream speed report](https://s3-public-assets.rpcfast.com/streaming-speed-comparison.html) was generated on **August 25, 2026** from a 3-minute local receive-time comparison of TxStream, ShredStream, and Yellowstone. The benchmark matched identical transaction signatures and used the account-include filter `pAMMBay6oceH9fJKBRHGP5D4bD4sWpmSwMn52FMfXEA`.
TxStream and ShredStream were nearly even on shared transactions (**49.1%** vs **50.9%** first-arrival share). Both shred-derived streams arrived before Yellowstone in more than **99.8%** of matched races.
You can track the streaming metrics comparison in our [Grafana dashboard](https://solana-metrics.rpcfast.com/public-dashboards/7ddaf20ee1864ecaba494e079a78343c).
**Geolocation:** Currently, we are present in Frankfurt, **Equinix FR13** data center.
{% hint style="warning" %}
For the most efficient performance and lowest latency, we recommend deploying your app in the **same region data center** or close. Recommended providers: Cherry Servers, Teraswitch, Latitude, AWS, GCP.
{% endhint %}
***
### Benchmarking Against Competitors
Use [GeyserBench](https://github.com/hmstudio-labs/geyserbench/blob/main/README.md) to compare multiple Solana gRPC endpoints simultaneously and measure their speed and reliability in detecting transactions. Compare latency percentile profiles and reliability for each RPC endpoint.
👉 **RPC Fast vs. Shyft** (Sept 1, 2026)
* [TxStream vs. Rabbitstream](https://s3-public-assets.rpcfast.com/benches/txstream-vs-rabbitstream.html)
* [Yellowstone gRPC comparison](https://s3-public-assets.rpcfast.com/benches/rpcfast-vs-shyft-beta.html) (with the Shyft's recent Beta version)
Note: 0.06ms ping between the servers compared.
👉 **RPC Fast vs. Triton** (May 13, 2026)
[Yellowstone gRPC comparison](https://runs.solstack.app/run/320734d7-3556-4934-8cab-2f7c72a49c06)
Note: The competitor's endpoint in this test was masked to avoid exposing the client's data.
***
### Benchmarking RPC Fast gRPC Services
#### TxStream vs. ShredStream vs. Yellowstone
[Open the full benchmark report](https://s3-public-assets.rpcfast.com/streaming-speed-comparison.html) (August 25, 2026)
The test measured local receive time for identical transaction signatures over 3 minutes. All endpoints used the same account-include filter.
| Faster endpoint in the report | Compared with | Shared transactions | Arrived first | Median lead when first | p95 lead when first |
| ----------------------------- | ------------- | ------------------: | ------------: | ---------------------: | ------------------: |
| TxStream | Yellowstone | 103,073 | 99.8% | 5.02 ms | 17.08 ms |
| ShredStream | Yellowstone | 58,688 | 99.9% | 4.82 ms | 16.04 ms |
| ShredStream | TxStream | 58,688 | 50.9% | 30 us | 571 us |
TxStream and ShredStream were effectively tied in the matched-signature comparison: TxStream arrived first in **49.1%** of shared transactions and ShredStream in **50.9%**.
| Endpoint | Unique transactions | Coverage | Overall first-arrival share |
| ----------- | ------------------: | -------: | --------------------------: |
| TxStream | 103,073 | 99.9% | 70.9% |
| ShredStream | 58,688 | 56.9% | 29.0% |
| Yellowstone | 103,177 | 100.0% | 0.1% |
The overall first-arrival share reflects both latency and which signatures each endpoint observed. Use the pairwise rows for a direct speed comparison because they include only signatures shared by both endpoints. Coverage is specific to this run and filter, not a general reliability guarantee.
{% hint style="info" %}
All endpoints were timestamped at the receive event on the same observer. See the full report for configured endpoints, batch-position analysis, and raw JSON.
{% endhint %}
***
## WebSockets Performance
According to independent [benchmark results](https://x.com/CompareNodes/status/2077002160659677359?s=20) by CompareNodes (as of July 14), RPC Fast WebSockets deliver messages about Solana slots **faster than Alchemy.**
Measured from 30 locations including Frankfurt using CompareNodes' public RPC Inspector Pro.
***
## Transaction Landing Performance: Beam
{% hint style="info" %}
[Benchmark code on Github](https://github.com/dysnix/solana-test/tree/main/sendtx-bench-rs). Create the type of transaction you need and check real-world performance.
{% endhint %}
### [Beam](https://docs.rpcfast.com/rpc-fast-saas-solana/sending-transactions/rpc-fast-beam-beta) Providers Performance Scoring
We evaluate each provider across five key performance metrics:
* Landing latency (milliseconds)
* Landing latency (slots)
* Transaction position within the block
* Same-slot landing rate
* Overall success rate
Each metric is normalized on a **0–100 scale** relative to the best-performing provider in the comparison set.
**For latency-related metrics,** p90 performance is weighted 4× more heavily than average performance, prioritizing consistency under load and minimizing tail latency.
**Block position** uses equal weighting between average and p90 values.
**The final performance score** is calculated as the unweighted average of all five metric scores.
#### **Five buckets:**
```
landed_ms = 0.2 × avg_score(submit_to_landed_grpc_ms)
+ 0.8 × p90_score(submit_to_landed_grpc_ms)
```
```
landed_slots = 0.2 × avg_score(submit_to_landed_slots)
+ 0.8 × p90_score(submit_to_landed_slots)
```
```
landed_idx = 0.5 × avg_score(landed_index_in_block)
+ 0.5 × p90_score(landed_index_in_block)
```
```
same_slot = higher_is_better_score(same_slot_landed_count)
```
```
success_ratio = higher_is_better_score(landed_runs / total_runs)
```
#### **Final score:**
```
performance_rate_pct = (landed_ms + landed_slots + landed_idx + same_slot + success_ratio) / 5
```
***
### Beam Providers Comparison
* Generated: `2026-05-19T15:16:33.756079549+00:00`
* Runs per provider: `10`
* Selected providers: `beam-astralane, beam-bloxroute, beam-falcon`
| Provider | Avg Ack ms | P90 Ack ms | Avg Landed ms | P90 Landed ms | Avg Block ms | P90 Block ms | Avg Landed slots | P90 Landed slots | Avg Idx-in-Block | P90 Idx-in-Block | Avg Priority Fee | Max Slots | Min Slots | Same-slot landed | Landed runs | Block seen | Total runs | Success ratio % | Performance rate % |
| ------------------ | ---------- | ---------- | ------------- | ------------- | ------------ | ------------ | ---------------- | ---------------- | ---------------- | ---------------- | ---------------- | --------- | --------- | ---------------- | ----------- | ---------- | ---------- | --------------- | ------------------ |
| **beam-astralane** | 51.793 | 54.891 | 231.625 | 443.929 | 393.336 | 453.750 | 0.700 | 1.000 | 526.30 | 773.00 | 10000.000 | 1 | 0 | 3 | 10 | 10 | 10 | 100.00 | 99.78 |
| **beam-bloxroute** | 53.163 | 54.728 | 232.784 | 443.948 | 391.469 | 453.708 | 0.700 | 1.000 | 655.50 | 829.00 | 10000.000 | 1 | 0 | 3 | 10 | 10 | 10 | 100.00 | 99.66 |
| **beam-falcon** | 53.888 | 54.612 | 229.601 | 443.934 | 390.611 | 453.725 | 0.700 | 1.000 | 658.30 | 830.00 | 10000.000 | 1 | 0 | 3 | 10 | 10 | 10 | 100.00 | 100.00 |
**Note:** The benchmark is optimized for same-slot landing. A streaming slot signal triggers transaction submission immediately as a new slot begins, and landing confirmation is determined from multiple data sources. Yellowstone gRPC is the default signal source unless another stream is explicitly selected.
***
### Beam vs RPC Comparison
* Generated: `2026-05-19T15:22:23.427708898+00:00`
* Runs per provider: `20`
* Selected providers: `beam-astralane, rpcfast-rpc`
| Provider | Avg Ack ms | P90 Ack ms | Avg Landed ms | P90 Landed ms | Avg Block ms | P90 Block ms | Avg Landed slots | P90 Landed slots | Avg Idx-in-Block | P90 Idx-in-Block | Avg Priority Fee | Max Slots | Min Slots | Same-slot landed | Landed runs | Block seen | Total runs | Success ratio % | Performance rate % |
| ------------------ | ---------- | ---------- | ------------- | ------------- | ------------ | ------------ | ---------------- | ---------------- | ---------------- | ---------------- | ---------------- | --------- | --------- | ---------------- | ----------- | ---------- | ---------- | --------------- | ------------------ |
| **beam-astralane** | 52.586 | 55.406 | 330.392 | 589.025 | 345.213 | 589.025 | 0.400 | 1.000 | 704.45 | 1092.00 | 10000.000 | 1 | 0 | 12 | 20 | 20 | 20 | 100.00 | 97.42 |
| **rpcfast-rpc** | 27.570 | 38.891 | 457.492 | 1065.526 | 503.728 | 1065.526 | 0.650 | 2.000 | 574.65 | 1012.00 | 10000.000 | 4 | 0 | 12 | 20 | 20 | 20 | 100.00 | 82.20 |
***
## RPC Fast Endpoint Checker
Quick health checker for your Solana endpoint (mainnet):
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:
```
GET https://docs.rpcfast.com/rpc-fast-saas-solana/performance-and-benchmarking.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
