Исследование с проверкой источников: Coinbase Websocket Market DATA

Verified source facts

1. Coinbase says revised WebSocket market-data subscription pricing takes effect on September 15, 2026. 2. The default tier remains 10 subscriptions per product and channel at no monthly charge, and attempts above 10 are rejected without a paid tier. 3. The September 15 monthly prices are $300 for 15, $1,200 for 25, $2,400 for 50, $4,200 for 100, $7,800 for 250 and $12,000 for 500 subscriptions. 4. Coinbase says one subscription is measured per product and channel and allows multiple products and channels to be included in one subscription request.

Comparison at a glance

1. Official-product check: Effective time

A venue announcement becomes useful only after its rule is translated into a testable account decision. Here, effective time concerns September 15, 2026. Retain the dated first-party wording, identify the eligible entity and product screen, and then record the account billing boundary. A launch headline cannot prove access, suitability, liquidity, or the result for a particular account.

Separate the product's stated mechanics from promotional framing. Identify the funding source, matching or pricing rule, settlement trigger, fee formula, loss boundary, and suspension policy. The documentation can describe those controls, but it cannot resolve timezone presentation. That unknown must remain visible rather than being converted into a prediction.

Finally, make effective time reversible where possible: record the account billing boundary. Retain non-secret evidence and reconcile the result with the rule for September 15, 2026. This section evaluates operational mechanics; because timezone presentation remains unresolved, it does not predict the underlying market or recommend taking the exposure.

2. Official-product check: Default allowance

Here, default allowance concerns 10 per product and channel. Retain the dated first-party wording, identify the eligible entity and product screen, and then measure the same unit.

The documentation can describe those controls, but it cannot resolve account exceptions.

Finally, make default allowance reversible where possible: measure the same unit. Retain non-secret evidence and reconcile the result with the rule for 10 per product and channel. This section evaluates operational mechanics; because account exceptions remains unresolved, it does not predict the underlying market or recommend taking the exposure.

3. Official-product check: Current usage

Here, current usage concerns live product-channel subscriptions. Retain the dated first-party wording, identify the eligible entity and product screen, and then inventory before selecting a tier.

The documentation can describe those controls, but it cannot resolve duplicate clients.

Finally, make current usage reversible where possible: inventory before selecting a tier. Retain non-secret evidence and reconcile the result with the rule for live product-channel subscriptions. This section evaluates operational mechanics; because duplicate clients remains unresolved, it does not predict the underlying market or recommend taking the exposure.

4. Official-product check: Duplicate demand

Here, duplicate demand concerns repeated subscriptions to the same pair. Retain the dated first-party wording, identify the eligible entity and product screen, and then remove unnecessary duplication.

The documentation can describe those controls, but it cannot resolve intentional redundancy.

Finally, make duplicate demand reversible where possible: remove unnecessary duplication. Retain non-secret evidence and reconcile the result with the rule for repeated subscriptions to the same pair. This section evaluates operational mechanics; because intentional redundancy remains unresolved, it does not predict the underlying market or recommend taking the exposure.

5. Official-product check: Tier cost

Here, tier cost concerns monthly capacity price. Retain the dated first-party wording, identify the eligible entity and product screen, and then calculate required headroom.

The documentation can describe those controls, but it cannot resolve future repricing.

Finally, make tier cost reversible where possible: calculate required headroom. Retain non-secret evidence and reconcile the result with the rule for monthly capacity price. This section evaluates operational mechanics; because future repricing remains unresolved, it does not predict the underlying market or recommend taking the exposure.

6. Official-product check: Request batching

Here, request batching concerns multiple products and channels per request. Retain the dated first-party wording, identify the eligible entity and product screen, and then test reconnect behavior.

The documentation can describe those controls, but it cannot resolve message-size limit.

Finally, make request batching reversible where possible: test reconnect behavior. Retain non-secret evidence and reconcile the result with the rule for multiple products and channels per request. This section evaluates operational mechanics; because message-size limit remains unresolved, it does not predict the underlying market or recommend taking the exposure.

7. Official-product check: Inbound rate

Here, inbound rate concerns subscription request token bucket. Retain the dated first-party wording, identify the eligible entity and product screen, and then pace automated reconnects.

The documentation can describe those controls, but it cannot resolve redeploy burst.

Finally, make inbound rate reversible where possible: pace automated reconnects. Retain non-secret evidence and reconcile the result with the rule for subscription request token bucket. This section evaluates operational mechanics; because redeploy burst remains unresolved, it does not predict the underlying market or recommend taking the exposure.

8. Official-product check: Feed choice

Here, feed choice concerns full, level2 or batch channel. Retain the dated first-party wording, identify the eligible entity and product screen, and then choose minimum necessary verbosity.

The documentation can describe those controls, but it cannot resolve strategy requirements.

Finally, make feed choice reversible where possible: choose minimum necessary verbosity. Retain non-secret evidence and reconcile the result with the rule for full, level2 or batch channel. This section evaluates operational mechanics; because strategy requirements remains unresolved, it does not predict the underlying market or recommend taking the exposure.

9. Official-product check: Failover

Here, failover concerns secondary connection and recovery. Retain the dated first-party wording, identify the eligible entity and product screen, and then exercise the switchover.

The documentation can describe those controls, but it cannot resolve shared quota.

Finally, make failover reversible where possible: exercise the switchover. Retain non-secret evidence and reconcile the result with the rule for secondary connection and recovery. This section evaluates operational mechanics; because shared quota remains unresolved, it does not predict the underlying market or recommend taking the exposure.

10. Official-product check: Billing evidence

Here, billing evidence concerns selected tier and invoices. Retain the dated first-party wording, identify the eligible entity and product screen, and then retain non-secret records.

The documentation can describe those controls, but it cannot resolve account-specific taxes.

Finally, make billing evidence reversible where possible: retain non-secret records. Retain non-secret evidence and reconcile the result with the rule for selected tier and invoices. This section evaluates operational mechanics; because account-specific taxes remains unresolved, it does not predict the underlying market or recommend taking the exposure.

Decision scenarios

Scenario 1: when effective time becomes decisive

Consider a reader whose main constraint is September 15, 2026. The correct response is not to infer a universal platform or product winner. It is to record the account billing boundary, retain the evidence, and define a stop condition before funds or permissions are committed. If the evidence changes, the decision must be reopened. This produces a repeatable operating rule rather than a one-time opinion.

Scenario 2: when default allowance becomes decisive

Consider a reader whose main constraint is 10 per product and channel. It is to measure the same unit, retain the evidence, and define a stop condition before funds or permissions are committed.

Scenario 3: when current usage becomes decisive

Consider a reader whose main constraint is live product-channel subscriptions. It is to inventory before selecting a tier, retain the evidence, and define a stop condition before funds or permissions are committed.

Scenario 4: when duplicate demand becomes decisive

Consider a reader whose main constraint is repeated subscriptions to the same pair. It is to remove unnecessary duplication, retain the evidence, and define a stop condition before funds or permissions are committed.

Scenario 5: when tier cost becomes decisive

Consider a reader whose main constraint is monthly capacity price. It is to calculate required headroom, retain the evidence, and define a stop condition before funds or permissions are committed.

Scenario 6: when request batching becomes decisive

Consider a reader whose main constraint is multiple products and channels per request. It is to test reconnect behavior, retain the evidence, and define a stop condition before funds or permissions are committed.

Sources and editorial method

• Coinbase Market Data Connections
• Coinbase Exchange WebSocket Rate Limits
• Coinbase Exchange WebSocket Best Practices