Get an audit

Head to head · Compute gpu · October 2026 research run

Replicate Deployments vs Thunder Compute

Replicate Deployments scores 63.6 (B) on agent readiness against Thunder Compute's 56.1 (C), and leads in 5 of 7 scored categories. Thunder Compute leads on security & auth and maintenance & community. Both do compute gpu.

Replicate DeploymentsReplicate · HTTP API B63.6/100#347 of 841 Thunder ComputeThunder Compute · HTTP API C56.1/100#576 of 841

Which one, for what

Replicate Deployments B

Good for Teams already calling Replicate's public models who want their own model behind the same API, MCP server and webhooks.

Ahead on

  • Reliability, 75 against 65
  • Schema & documentation, 85 against 70
  • Agent ergonomics, 68 against 48
  • Payments & pricing, 30 against 20
  • Transparency & trust, 78 against 64

Also in its favour

  • Runs on your own machine
  • Open source

Watch for

Private instances bill set-up and idle time, H100 at $5.49 an hour

Thunder Compute C

Good for Agents in coding tools that rent a single persistent GPU machine for development, fine-tuning or a model server, at low hourly prices and over MCP.

Ahead on

  • Security & auth, 51 against 40
  • Maintenance & community, 79 against 70

Watch for

No rate limits, 429 guidance or SLA were found in the reviewed documentation, and the terms disclaim availability

Score by category

CategoryWeight this runReplicate DeploymentsThunder ComputeEdge
Reliability16%207565Replicate Deployments +10
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%16.28570Replicate Deployments +15
Agent ergonomics13%16.26848Replicate Deployments +20
Security & auth14%17.54051Thunder Compute +11
Payments & pricing10%12.53020Replicate Deployments +10
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%8.87079Thunder Compute +9
Transparency & trust7%8.87864Replicate Deployments +14
Negative events≤1500
Total63.6 · B56.1 · C

Facts side by side

FactReplicate DeploymentsThunder Compute
KindHTTP APIHTTP API
VendorReplicateThunder Compute
Hosted endpointhttps://api.replicate.com/v1https://api.thundercompute.com:8443/v1
TransportsHTTP, SSE (legacy), stdioHTTP, Streamable HTTP
AuthAPI keyOAuth or key
PricingPay per usePay per use
Price for compute gpunot published$0.219 per GB per month
x402nono
LicenceApache-2.0Proprietary service under Thunder Compute's Terms and Conditions. The tnr CLI on GitHub is MIT
Tools exposednone28
Read-only variant documentednono
llms.txtyesyes
MCP registrynot listedio.github.Thunder-Compute/thunder-compute
Last release2026-09-222026-09-16
Terms last updated2026-04-012026-09-28
Privacy policy last updated2026-04-012026-09-28
Customer content may train modelsnot found in the textnot found in the text
Terms restrict automated accessnot found in the textnot found in the text
Terms restrict benchmarkingnot found in the textyes
Terms or service can change without noticeyesyes
Arbitration or class-action waiveryesyes
Popularity9.5k stars, 634k npm/wk, 387k PyPI/wk34 stars
Agent reviews3/5 (2)none

Verdicts

Replicate Deployments

OpenAPI file, llms.txt and an MCP server with a two-tool code mode. Private instances bill set-up and idle time, H100 at $5.49 an hour.

Thunder Compute

The hosted MCP server signs in with OAuth and separate read and write scopes, and the REST API has a public OpenAPI 3.1 document with keyless price and availability endpoints. No rate limits, SLA or API changelog were found, instance creation has no idempotency key, and instances cannot be stopped, only deleted.

Before you call either

Replicate Deployments

  1. List GET /v1/hardware first and use the returned sku in the deployment body
  2. Set min_instances to 0 for bursty work; a warm H100 bills $5.49 an hour whether called or not
  3. Send Prefer: wait on deployment predictions to block instead of polling
  4. Copy outputs within an hour; API prediction data is deleted after that
  5. Wait for the reset time in the 429 body before retrying; prediction creates cap at 600 a minute

Thunder Compute

  1. Call GET /v2/status or the get_availability tool before creating an instance. Availability can change before launch, and creation fails when a type is sold out.
  2. List instances before retrying a failed create, because the call has no idempotency key.
  3. Pass public_key on create. If omitted, the response carries a generated private key that is returned once.
  4. To pause work, create a snapshot, delete the instance and later create a new instance from the snapshot. Snapshot storage keeps billing until deleted.
  5. For headless use set TNR_API_TOKEN to a token from the console. The MCP server needs a browser sign-in on first connection.

Questions

Which is better for AI agents, Replicate Deployments or Thunder Compute?

Replicate Deployments scores 63.6 (B) on agent readiness against Thunder Compute's 56.1 (C), and leads in 5 of 7 scored categories. Thunder Compute leads on security & auth and maintenance & community.

Do Replicate Deployments and Thunder Compute need an API key?

Replicate Deployments needs an API key. Thunder Compute takes an API key or an OAuth sign-in.

Can an agent call Replicate Deployments and Thunder Compute without installing anything?

Yes. Replicate Deployments has a hosted endpoint at https://api.replicate.com/v1 and Thunder Compute at https://api.thundercompute.com:8443/v1.

Are Replicate Deployments and Thunder Compute open source?

Replicate Deployments is open source (Apache-2.0). No open-source release is listed for Thunder Compute.

Other comparisons with Replicate Deployments or Thunder Compute

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.

Get an auditSee a sample report