MeshedFlow
Pricing one-pager

Pricing follows verified bandwidth you stop buying.

Start with one measured stream. If MeshedFlow does not show useful offload with acceptable playback health, there is no rollout contract to defend.

Review proof
Commercial shape

One pilot, then a commercial model built from your traffic.

The first quote is not a seat count or a generic platform tax. It is a measured business case from control traffic, treatment traffic, CDN fallback, and verified peer-delivered bytes.

01 Measured pilot

No upfront software license for one stream or event window. Includes baseline, treatment, playback health, fallback reasons, and finance-readable report.

02 Canary expansion

Limited paid expansion when the report shows useful offload. Scope stays constrained by stream, geography, device class, or traffic share.

03 Production rollout

Commercial terms use verified offloaded traffic, support requirements, and operational controls rather than a flat viewer tax.

Pricing model

The pilot turns CDN waste into a payable unit.

The commercial unit is verified offloaded GB or TB. The buyer can compare it directly against their CDN egress rate and event economics.

Measured offloadpeer-delivered bytes
Gross avoided spendoffloaded GB x CDN $/GB
Expansion ruleonly quote after pilot evidence
One event example

$100K used to be the bill. Now the target is $60K.

Plain version: offload half the CDN traffic, share a small slice of the savings, and the customer keeps the rest.

Before $100K

standard event bill

Now $60K

final delivery bill

40% reduction
Original bill $100K
New bill $60K
50%

$100K was the original bill. On this event, MeshedFlow offloads 50% of CDN traffic to the P2P network. That cuts CDN usage by $50K. MeshedFlow takes up to 20% of the saved bill: 20% of $50K = $10K. Final CDN + MeshedFlow bill: $60K. You only pay for what we save.

Included in pilot

Enough evidence for engineering and finance to say yes or no.

The pilot is designed to produce a decision, not a science project.

Baseline CDN-only bytes, viewer count, playback health, and current event economics.

Treatment P2P offloaded bytes, fallback bytes, connection success, routing reasons, and validation failures.

Safety Remote kill switch, upload caps, CDN fallback behavior, and rollback timing.

Decision Pass/fail, risks found, required fixes, next cohort, and commercial expansion recommendation.

Evidence boundary

What can be said before a customer pilot.

These are the latest measured rows. They support a pilot ask; they do not replace customer-specific proof.

Use in sales copy Evidence class Boundary
Latest 256-browser observe-vs-score E2E row represented all 256 peers: observe reached 85.00% total offload and 84.57% steady offload with failed media 0. Browser media proof, 2026-07-07 Current production decision row; production remains in observe and customer savings still require a pilot.
In the same 256-browser row, score reached 77.43% total offload and 77.63% steady offload, but had higher WSS/WebRTC churn than observe. Canary proof, 2026-07-07 Score stays canary-only. The small-swarm proof was promising, but the 256-peer production confirmation favored observe.
1024-browser E2E row represented all 1024 peers with 61.68% observe-mode offload, 69.37% score-mode offload, 79.79% at 512 browsers, and 0 failed media requests. Large browser proof, 2026-07-05 Large-row offload evidence only; the newer 256 decision row controls the production posture.
The low-density 32-peer E2E rerun showed score-mode offload 70.01%, join-after-WSS-open p95/p99 0.107s / 0.119s, first-peer p95 0.114s, and failed media 0. Timing proof, 2026-07-07 Use for score canary sanity only; the 256 decision row showed observe is still the safer production default.
Per-media browser rows remain green across HLS, protected media, live-edge, and ClearKey playback. Browser media proof, 2026-07-02 Customer production offload must be measured on their stream, devices, regions, and concurrency.
Hosted scale test passed 300K sessions with full cleanup. Hosted scale proof, 2026-07-02 This is service scale evidence, not a claim of 300K real browsers.
The latest ceiling probe found a 375K band, leaving roughly 25% instance headroom above the 300K operating row. Ceiling probe, 2026-07-02 Useful capacity data, not a 400K pass or a customer guarantee.
Next step

Bring one event, one CDN bill, and one release owner.

We will scope the stream, confirm the safety controls, run the comparison, and hand back a report that says whether expansion is worth buying.

Back to main site