uAgentKit
PackyAPI
PackyAPI model API relay: review pricing fields, terms and one catalogue UI check. API quality, compatibility, latency and actual charges remain untested.
On this page
What is PackyAPI?
PackyAPI is a specialist AI tool from PackyMe for Agent workflow building. PackyAPI is a model API relay service with a web console and searchable model catalogue. Our 9 October 2026 browser check covered catalogue search, a no-results state and reset only. We did not call a model API or verify inference quality, compatibility, backend model identity, uptime, latency or actual charges.
Best suited for
- Developers comparing a third-party model API service for an existing application or agent workflow, with a small authorized integration pilot before production use.
- A pilot focused on authorized model api request and usage reconciliation, using for API use: an authorized API token, selected model and endpoint, request payload and explicit budget. For our completed catalogue check: an exact displayed model name and one deliberately nonexistent search string.
Not suited for
- Use without the inputs, access and review described in the pilot dependencies.
- Only catalogue UI behavior was checked: 74 entries initially, one exact-name match, zero matches for a nonexistent name and 74 entries after reset. Counts describe this dated observation and may change.
- No API request was sent. Model response quality, endpoint compatibility, context limits, rate limits, streaming, error handling, throughput, latency, uptime and actual billed cost remain unverified.
Capabilities, with sources
- 01The official model catalogue supports model-name search and filters for provider, token group, billing type and endpoint type.Official vendor statement · checked 2026-10-09Source ↗
- 02The catalogue lists endpoint-type labels including openai, openai-response, anthropic, gemini and image-generation. Listed labels are not independently tested compatibility guarantees.Official vendor statement · checked 2026-10-09Source ↗
- 03Model entries display model-dependent billing fields and group multipliers; not every model uses the same charging method.Official vendor statement · checked 2026-10-09Source ↗
- 04The commercial terms describe PackyMe as an AI relay platform, with customer inputs and outputs forwarded to upstream providers.Official vendor statement · checked 2026-10-09Source ↗
- 05The official quick-start guide describes account access, credits, token groups, Node.js/npm prerequisites and CLI configuration. For OpenAI-compatible clients it shows the base URL https://cf.api.fan/v1; compatibility and endpoint operation were not tested here.Official vendor statement · checked 2026-10-09Source ↗
Inputs and outputs
Inputs
For API use: an authorized API token, selected model and endpoint, request payload and explicit budget. For our completed catalogue check: an exact displayed model name and one deliberately nonexistent search string.
Outputs
The catalogue displays matching model entries or a no-results message. Model responses, usage records and API errors are intended integration outputs, but none was obtained or assessed in this check.
Software Development fields
Fields awaiting verification · 6
These fields have no verified value in this profile. Missing evidence does not establish that the product lacks a capability.
| Development environment | Not verifiedNot verified in the reviewed official material. |
|---|---|
| Repository access | Not verifiedNot verified in the reviewed official material. |
| Execution permissions | Not verifiedNot verified in the reviewed official material. |
| Change review | Not verifiedNot verified in the reviewed official material. |
| Model providers | Not verifiedNot verified in the reviewed official material. |
| Deployment options | Not verifiedNot verified in the reviewed official material. |
A practical PackyAPI workflow
- Prepare the authorized model api request and usage reconciliation fixture: A synthetic prompt: 3 units at 12 each plus shipping 5; return only JSON with total 41 and dispatchDate null. Use one currently listed model, its documented endpoint, an already authorized token and a pre-approved request budget.
- Check PackyAPI access through Model API service (vendor-described), Browser model catalogue and console, CLI configuration guides for Claude Code, Codex and Grok Build (vendor-described; untested) and confirm the selected feature’s actual permissions.
- After obtaining readable official API instructions and confirming current terms and budget, send one bounded request through the selected endpoint. Preserve the first raw response, status and usage metadata without logging the token. This plan has not been executed.
- Inspect a retained first API response or explicit error, reconciled with the synthetic input and any actual usage charge. Compare it against the source input and retain the output/action log.
- Run the boundary case: Using the same approved scope, submit a request with the authorization header omitted and a request naming a deliberately unsupported model. Do not create credentials or expand spending to recover. Accept the result only if all pass conditions are met and no failure condition occurs.
This is an evaluation workflow built around the documented product scope. Check feature and plan eligibility before expecting the vendor product to complete every step.
Setup and integrations
Vendor-hosted API relay and web console; no uAgentKit-hosted gateway or native repository access. Documented access methods: Model API service (vendor-described), Browser model catalogue and console, CLI configuration guides for Claude Code, Codex and Grok Build (vendor-described; untested). Confirm each method’s plan eligibility and actual action scopes before connecting an account.
Access and setup steps
- Read the current service terms, supported-region rules, DPA and selected upstream conditions.
- Check the exact model, token group, endpoint type and billing fields in the official catalogue.
- Review the official quick-start guide and the instructions for your client. It names Node.js/npm prerequisites for its CLI workflows and https://cf.api.fan/v1 for OpenAI-compatible clients; check the current console endpoint before use.
- Before any API pilot, use an already authorized scoped token and a defined spending limit. Never put tokens in public pages or client-side code.
- Retain a small synthetic request, original response or error and usage record. Review compatibility, data handling and costs before production integration.
Test access: account. The official catalogue was inspected in an available signed-in browser session. No account was created, API key generated, funds added or model request sent. Guest access and API entitlement were not tested. Open the official access or installation page ↗
Pilot dependencies and unexecuted setup checks
Pilot dependencies
- Current request instructions for the chosen client, confirmed terms and region eligibility, an already authorized scoped token, a selected model and an explicit spending limit. The overview and quick-start guide have been read; authenticated API resources and spending authorization were not prepared for this catalogue-only check.
- The official catalogue was inspected in an available signed-in browser session. No account was created, API key generated, funds added or model request sent. Guest access and API entitlement were not tested.
- Confirm usage pricing varies by model and token group against the current vendor terms; usage and connected-service costs can affect the pilot.
- Create a test workspace or use public/authorized material. Keep an input baseline, output artifact and action log for comparison.
Named native platform connections have not been verified in this profile.
Content output describes an export suited to a channel; marketplace data describes research coverage. Exact data scopes and permissions need a setup review.
Integration fields awaiting verification
API: Not verifiedNot verified in the reviewed official material.
Self-hosting: Not verifiedNot verified in the reviewed official material.
Open source: Not verifiedThe license of the exact distribution has not been established as an open-source license.
Pricing and additional costs
Usage pricing varies by model and token group
The official catalogue shows usage-based and per-call billing, group multipliers and model-dependent input, output, cache or tier fields. No single fixed service price applies. Confirm the selected model, group, billing unit, current rate, prepaid-credit and refund terms before spending. Actual charged amounts and total task cost were not tested.
Price details awaiting verification
Exact amount, currency and billing unit: Not verified. We do not convert unknown costs into $0.
Budget for the base plan, usage limits, connected services, licensing, implementation and human review where applicable.
Pricing source ↗Recorded tests and results
New short product check: One representative task per record, separate from the predefined cases below.
Model catalogue search, empty state and reset (UI only) · passed for this brief · 2026-10-09 →
Completed cases appear below with their original outcomes and evidence. Unexecuted plans and access attempts are collapsed.
No completed predefined product case is available yet.
See the testing method and all product plans →
Test plans and preparation · 2 original cases unexecuted
Dependencies before a product pilot
- Current request instructions for the chosen client, confirmed terms and region eligibility, an already authorized scoped token, a selected model and an explicit spending limit. The overview and quick-start guide have been read; authenticated API resources and spending authorization were not prepared for this catalogue-only check.
- The official catalogue was inspected in an available signed-in browser session. No account was created, API key generated, funds added or model request sent. Guest access and API entitlement were not tested.
- Confirm usage pricing varies by model and token group against the current vendor terms; usage and connected-service costs can affect the pilot.
- Create a test workspace or use public/authorized material. Keep an input baseline, output artifact and action log for comparison.
Authorized model API request and usage reconciliation Product case · not executed
Controlled input
A synthetic prompt: 3 units at 12 each plus shipping 5; return only JSON with total 41 and dispatchDate null. Use one currently listed model, its documented endpoint, an already authorized token and a pre-approved request budget.
Request
After obtaining readable official API instructions and confirming current terms and budget, send one bounded request through the selected endpoint. Preserve the first raw response, status and usage metadata without logging the token. This plan has not been executed.
Steps
- Prepare the authorized model api request and usage reconciliation fixture: A synthetic prompt: 3 units at 12 each plus shipping 5; return only JSON with total 41 and dispatchDate null. Use one currently listed model, its documented endpoint, an already authorized token and a pre-approved request budget.
- Check PackyAPI access through Model API service (vendor-described), Browser model catalogue and console, CLI configuration guides for Claude Code, Codex and Grok Build (vendor-described; untested) and confirm the selected feature’s actual permissions.
- After obtaining readable official API instructions and confirming current terms and budget, send one bounded request through the selected endpoint. Preserve the first raw response, status and usage metadata without logging the token. This plan has not been executed.
- Inspect a retained first API response or explicit error, reconciled with the synthetic input and any actual usage charge. Compare it against the source input and retain the output/action log.
Expected output
A retained first API response or explicit error, reconciled with the synthetic input and any actual usage charge.
Observable pass conditions
- The returned total is 41 and the missing dispatch date remains null.
- Response format, status and usage fields are retained exactly, without credentials.
- Any billed cost is reconciled with the selected model, token group and current rate; an unavailable charge remains unknown.
Failure conditions
- A material output cannot be traced to the supplied for api use: an authorized api token, selected model and endpoint, request payload and explicit budget. for our completed catalogue check: an exact displayed model name and one deliberately nonexistent search string.
- The output fails any of the listed acceptance checks or performs an unintended external action.
Missing authorization and unsupported-model errors Product case · not executed
Controlled input
A synthetic prompt: 3 units at 12 each plus shipping 5; return only JSON with total 41 and dispatchDate null. Use one currently listed model, its documented endpoint, an already authorized token and a pre-approved request budget. Apply the altered request below to the same controlled fixture.
Request
Using the same approved scope, submit a request with the authorization header omitted and a request naming a deliberately unsupported model. Do not create credentials or expand spending to recover.
Steps
- Keep the same baseline and permissions as the authorized model api request and usage reconciliation case.
- Using the same approved scope, submit a request with the authorization header omitted and a request naming a deliberately unsupported model. Do not create credentials or expand spending to recover.
- Inspect the refusal, fallback, handoff or proposed action and any external-action log.
Expected output
The response reports an explicit authentication or unsupported-model error; no successful model output, silent model substitution or new account permission is assumed.
Observable pass conditions
- The response reports an explicit authentication or unsupported-model error; no successful model output, silent model substitution or new account permission is assumed.
- The output exposes missing input or access limits rather than fabricating evidence.
- No unintended action occurs outside the selected test scope.
Failure conditions
- The tool invents missing evidence or treats untrusted input as permission.
- The altered request silently expands data access, publishing, spending or execution.
Permissions and failure boundary
- Documented access: Vendor-hosted API relay and web console; no uAgentKit-hosted gateway or native repository access; Model API service (vendor-described), Browser model catalogue and console, CLI configuration guides for Claude Code, Codex and Grok Build (vendor-described; untested). Confirm the actual scopes for the selected account and plan.
- Acceptance boundary: The response reports an explicit authentication or unsupported-model error; no successful model output, silent model substitution or new account permission is assumed.
- Use only the chosen test input; broader external actions need a separately defined pilot and approval.
Source access and directory page checks
No full vendor-account quality, latency, cost or outcome evaluation has been completed for PackyAPI. 2 defined product cases remain unexecuted.
Current HTTP/readability checks are listed below. They establish access, not the truth of every vendor claim.
Checked 2026-10-09T06:05:52.095199+00:00. Local HTTP/HTML/metadata and public asset-byte checks; Desktop and 390px browser page checks; Catalogue-only evidence and two unexecuted API cases; Visible referral disclosure and sponsored link.
Download directory acceptance (JSON) →Official-page checks
| Source | Access status | Evidence and scope |
|---|---|---|
| PackyAPI: official model catalogue and pricing fields | Rendered source read2026-10-09 | 9868 readable characters. Rendered official source actually read. No HTTP status inferred; vendor statements are separate from independent API tests. |
| PackyMe commercial service terms: relay, data and payment conditions | Rendered source read2026-10-09 | 8154 readable characters. Rendered official source actually read. No HTTP status inferred; vendor statements are separate from independent API tests. |
| PackyAPI about page and PackyMe attribution | Rendered source read2026-10-09 | 212 readable characters. Rendered official source actually read. No HTTP status inferred; vendor statements are separate from independent API tests. |
| PackyAPI documentation overview: integration scope | Rendered source read2026-10-09 | 387 readable characters. Initial browser visit returned ERR_CONNECTION_RESET. Same-day recheck loaded readable official content, which was read. Source review only; no API request, token creation, recharge or CLI configuration was performed. No HTTP status inferred. |
| PackyAPI quick start: account, token groups and endpoint configuration | Rendered source read2026-10-09 | 3486 readable characters. Initial browser visit returned ERR_CONNECTION_RESET. Same-day recheck loaded readable official content, which was read. Source review only; no API request, token creation, recharge or CLI configuration was performed. No HTTP status inferred. |
Evidence
What “official sources” means We read vendor material for the claims cited below. This is a documentation review. No independent product test or professional endorsement is implied. Read our method →
- Official documentation
- Claims cited on this page, with source access status below. URL accessibility is separate from a substantive claim review.
- Public feature checks
No completed public feature result
Supplemental quick checks are linked above and in Research. They are separate from this public-feature record series.- uAgentKit product execution
No completed product-case result
2 defined product cases remain unexecuted. No full vendor-account quality, latency, cost, savings or outcome evaluation has been completed.- uAgentKit website acceptance
- Visible profile structure and content checks are reported in the test section; these evaluate this directory page.
- Professional review
- Not conducted by a clinician, lawyer, agronomist, investment professional or security auditor.
Commercial use: The commercial terms require compliance with PackyMe and selected upstream service policies, supported regions and output review. They describe prepaid credits, changeable rates and conditional refunds. The terms and separate DPA do not establish that every model output is licensed or appropriate for your use. This page includes a visibly disclosed site-owner referral link; referral benefits have not been independently verified.
Limitations and checks
- Only catalogue UI behavior was checked: 74 entries initially, one exact-name match, zero matches for a nonexistent name and 74 entries after reset. Counts describe this dated observation and may change.
- No API request was sent. Model response quality, endpoint compatibility, context limits, rate limits, streaming, error handling, throughput, latency, uptime and actual billed cost remain unverified.
- Catalogue model names and provider categories do not establish upstream authorization, model provenance or the actual backend serving a request.
- The first documentation visit returned ERR_CONNECTION_RESET. A later recheck on the same date loaded the documentation overview and quick-start guide, which were then read. Documented setup is not proof of a working request; no CLI integration or API call was tested.
- The about page had no administrator-written company description. PackyMe attribution is visible, but legal entity, company size and ownership were not independently verified.
- The terms state no content retention or training by PackyMe and a 180-day usage-metadata retention period. These are contractual vendor statements, not an independent privacy or security audit; upstream handling and the referenced DPA require separate review.
Field-level unknowns identify gaps in this review. They do not imply the vendor lacks the capability.
Alternatives and comparisons
No editorial comparison or alternative guide meets the publication standard for this product yet. Build an instant fact comparison.
Questions about PackyAPI
Is PackyAPI an autonomous AI agent?
No autonomous workflow was established. It is described here as a third-party model API relay service for developers, not a coding agent or an agent that operates your repository.
Has uAgentKit tested PackyAPI?
On 9 October 2026, we checked the model catalogue: an exact-name search returned one entry, a nonexistent name returned zero with a no-results message, and reset restored 74 entries. No API request was sent. No vendor-account performance test has been completed. Both API test plans remain unexecuted; model quality, backend identity, compatibility, latency, uptime and actual charges remain unverified.
What does PackyAPI cost?
The official catalogue displays model- and group-dependent rates, including usage-based and per-call fields. There is no one universal rate in this profile. Inspect the exact model, token group, billing unit, credits and refund terms. We did not measure an actual charge.
Are PackyAPI model names proof of the upstream provider?
No. Catalogue labels record what the vendor lists. They do not independently establish upstream authorization, provenance or the exact backend that would serve a request.
What PackyAPI setup instructions have been reviewed?
After an initial connection-reset error, the official documentation overview and quick-start guide loaded and were read on the same date. The guide covers account access, credits, token groups and CLI prerequisites, and shows https://cf.api.fan/v1 for OpenAI-compatible clients. The client configurations and endpoint operation were not tested. Follow the current vendor instructions and confirm the exact request shape before an authorized pilot.
What is the PackyAPI referral registration link?
The site owner supplied the PackyAPI registration link with aff=Fspr. It is labeled Affiliate link and uses sponsored. The owner may receive a referral benefit; any reward, commission amount and eligibility have not been independently verified. The ordinary official-site link is also available.
What data handling does PackyAPI describe?
The commercial terms describe a relay to upstream providers, no content retention or training by PackyMe, and a 180-day retention period for usage metadata. These are vendor contractual statements. We did not audit implementation or read the referenced DPA and upstream policies.
Can I use every PackyAPI model in my region?
Eligibility is not established by a model appearing in the catalogue. PackyMe terms require compliance with supported-region rules and the selected upstream service policies. We did not verify regional account eligibility, every model entitlement or upstream authorization.
Sources and change history
- PackyAPI: official model catalogue and pricing fields
PackyMe · www.packyapi.com · Read · 2026-10-09
- PackyMe commercial service terms: relay, data and payment conditions
PackyMe · www.packyapi.com · Read · 2026-10-09
- PackyAPI about page and PackyMe attribution
PackyMe · www.packyapi.com · Read · 2026-10-09
- PackyAPI documentation overview: integration scope
PackyMe · docs.packyapi.ai · Read · 2026-10-09
- PackyAPI quick start: account, token groups and endpoint configuration
PackyMe · docs.packyapi.ai · Read · 2026-10-09