AIRanking

AI Ranking / Data snapshot

Better questions.
Better AI choices.

Compare AI API services with transparent scores, public evidence and practical buying guides. Build a shortlist, understand the gaps, then test what matters to you.

Page 01

The evidence, side by side

AI API providers 1–40 of 500
RankProvider & evidenceData scoreReported uptimeReported latencyModelsReviewsPoliciesView source
1ToTokens

ToTokens: Public-data score 74.9, with 100% weighted field coverage. Reported uptime: 99.1%; reported latency: 4.8 s.

Provider profile →
74.9Coverage 100%99.1%4.8 s334 / 5 (8)Refund: YesInvoice: Yes↗
2云渡

云渡: Public-data score 79.1, with 100% weighted field coverage. Reported uptime: 100%; reported latency: 3.4 s.

Provider profile →
79.1Coverage 100%100%3.4 s305 / 5 (4)Refund: YesInvoice: Yes↗
3速蹬

速蹬: Public-data score 75, with 85% weighted field coverage. Reported uptime: 100%; reported latency: 4.4 s.

Provider profile →
75Coverage 85%100%4.4 s22Not recordedRefund: YesInvoice: Yes↗
488API Token聚合平台

88API Token聚合平台: Public-data score 65.1, with 100% weighted field coverage. Reported uptime: 96.1%; reported latency: 5.6 s.

Provider profile →
65.1Coverage 100%96.1%5.6 s685 / 5 (2)Refund: NoInvoice: Yes↗
5hao.ai

hao.ai: Public-data score 52.5, with 100% weighted field coverage. Reported uptime: 76%; reported latency: 3.4 s.

Provider profile →
52.5Coverage 100%76%3.4 s195 / 5 (3)Refund: YesInvoice: Yes↗
6灵算

灵算: Public-data score 76.3, with 100% weighted field coverage. Reported uptime: 99.7%; reported latency: 5.4 s.

Provider profile →
76.3Coverage 100%99.7%5.4 s344.4 / 5 (14)Refund: YesInvoice: Yes↗
7灵珑AI

灵珑AI: Public-data score 74.7, with 85% weighted field coverage. Reported uptime: 100%; reported latency: 4.6 s.

Provider profile →
74.7Coverage 85%100%4.6 s44Not recordedRefund: YesInvoice: Yes↗
89527code

9527code: Public-data score 71.9, with 100% weighted field coverage. Reported uptime: 93.9%; reported latency: 5.7 s.

Provider profile →
71.9Coverage 100%93.9%5.7 s654.9 / 5 (60)Refund: YesInvoice: Yes↗
9SudoCode

SudoCode: Public-data score 62.9, with 100% weighted field coverage. Reported uptime: 88.9%; reported latency: 4.3 s.

Provider profile →
62.9Coverage 100%88.9%4.3 s244.8 / 5 (9)Refund: YesInvoice: Yes↗
10破晓

破晓: Public-data score 41.3, with 80% weighted field coverage. Reported uptime: 77.1%; reported latency: 6.9 s.

Provider profile →
41.3Coverage 80%77.1%6.9 s29Not recordedRefund: YesInvoice: Yes↗
11ToolCode

ToolCode: Public-data score 76.2, with 100% weighted field coverage. Reported uptime: 100%; reported latency: 4 s.

Provider profile →
76.2Coverage 100%100%4 s195 / 5 (1)Refund: YesInvoice: Yes↗
12Modelflare

Modelflare: Public-data score 70.4, with 100% weighted field coverage. Reported uptime: 97.6%; reported latency: 5.4 s.

Provider profile →
70.4Coverage 100%97.6%5.4 s473.3 / 5 (20)Refund: YesInvoice: Yes↗
13ZNB.ai

ZNB.ai: Public-data score 63.8, with 100% weighted field coverage. Reported uptime: 97.4%; reported latency: 7.8 s.

Provider profile →
63.8Coverage 100%97.4%7.8 s345 / 5 (1)Refund: YesInvoice: No↗
14BuyToken

BuyToken: Public-data score 63.8, with 100% weighted field coverage. Reported uptime: 90.6%; reported latency: 6 s.

Provider profile →
63.8Coverage 100%90.6%6 s305 / 5 (1)Refund: YesInvoice: Yes↗
15FluxToken

FluxToken: Public-data score 58.3, with 85% weighted field coverage. Reported uptime: 89.1%; reported latency: 6.6 s.

Provider profile →
58.3Coverage 85%89.1%6.6 s45Not recordedRefund: YesInvoice: Yes↗
16悟空API

悟空API: Public-data score 64.8, with 100% weighted field coverage. Reported uptime: 94%; reported latency: 6.1 s.

Provider profile →
64.8Coverage 100%94%6.1 s304.2 / 5 (9)Refund: YesInvoice: Yes↗
17Yomi API

Yomi API: Public-data score 54.8, with 85% weighted field coverage. Reported uptime: 87.1%; reported latency: 6.3 s.

Provider profile →
54.8Coverage 85%87.1%6.3 s45Not recordedRefund: YesInvoice: Yes↗
18OmniaKey

OmniaKey: Public-data score 65.8, with 85% weighted field coverage. Reported uptime: 92.6%; reported latency: 5.5 s.

Provider profile →
65.8Coverage 85%92.6%5.5 s244Not recordedRefund: YesInvoice: Yes↗
19ccgo

ccgo: Public-data score 72.2, with 100% weighted field coverage. Reported uptime: 98.7%; reported latency: 5.6 s.

Provider profile →
72.2Coverage 100%98.7%5.6 s305 / 5 (1)Refund: YesInvoice: Yes↗
20吉吉不降智AI

吉吉不降智AI: Public-data score 69.2, with 85% weighted field coverage. Reported uptime: 96.2%; reported latency: 6.1 s.

Provider profile →
69.2Coverage 85%96.2%6.1 s44Not recordedRefund: YesInvoice: Yes↗
21Toknex

Toknex: Public-data score 75.4, with 100% weighted field coverage. Reported uptime: 100%; reported latency: 4.8 s.

Provider profile →
75.4Coverage 100%100%4.8 s245 / 5 (4)Refund: YesInvoice: Yes↗
22词元玩家

词元玩家: Public-data score 73.6, with 80% weighted field coverage. Reported uptime: 100%; reported latency: 2.3 s.

Provider profile →
73.6Coverage 80%100%2.3 s7Not recordedRefund: Not recordedInvoice: Yes↗
23词易

词易: Public-data score 59.3, with 100% weighted field coverage. Reported uptime: 87.3%; reported latency: 5.9 s.

Provider profile →
59.3Coverage 100%87.3%5.9 s414.4 / 5 (7)Refund: YesInvoice: Yes↗
24PackyCode

PackyCode: Public-data score 68.8, with 85% weighted field coverage. Reported uptime: 98.3%; reported latency: 5.8 s.

Provider profile →
68.8Coverage 85%98.3%5.8 s653.5 / 5 (4)Refund: Not recordedInvoice: Not recorded↗
25Framefly

Framefly: Public-data score 67.2, with 100% weighted field coverage. Reported uptime: 99.2%; reported latency: 4.3 s.

Provider profile →
67.2Coverage 100%99.2%4.3 s115 / 5 (1)Refund: NoInvoice: Yes↗
26ModCon

ModCon: Public-data score 70.2, with 95% weighted field coverage. Reported uptime: 99.4%; reported latency: 5 s.

Provider profile →
70.2Coverage 95%99.4%5 s155 / 5 (3)Refund: YesInvoice: Not recorded↗
27iKunAPI

iKunAPI: Public-data score 69.6, with 85% weighted field coverage. Reported uptime: 99.5%; reported latency: 4.1 s.

Provider profile →
69.6Coverage 85%99.5%4.1 s5Not recordedRefund: YesInvoice: Yes↗
28sudocode.us

sudocode.us: Public-data score 74.4, with 85% weighted field coverage. Reported uptime: 99.3%; reported latency: 4.9 s.

Provider profile →
74.4Coverage 85%99.3%4.9 s61Not recordedRefund: YesInvoice: Yes↗
29SevnX

SevnX: Public-data score 70, with 95% weighted field coverage. Reported uptime: 99.7%; reported latency: 4.6 s.

Provider profile →
70Coverage 95%99.7%4.6 s125 / 5 (5)Refund: Not recordedInvoice: Yes↗
308sToken

8sToken: Public-data score 77, with 100% weighted field coverage. Reported uptime: 98.6%; reported latency: 4.7 s.

Provider profile →
77Coverage 100%98.6%4.7 s775 / 5 (9)Refund: YesInvoice: Yes↗
31云舟AI

云舟AI: Public-data score 67.9, with 85% weighted field coverage. Reported uptime: 100%; reported latency: 4.9 s.

Provider profile →
67.9Coverage 85%100%4.9 s3Not recordedRefund: YesInvoice: Yes↗
32RunAPI

RunAPI: Public-data score 73.5, with 85% weighted field coverage. Reported uptime: 97.7%; reported latency: 5.8 s.

Provider profile →
73.5Coverage 85%97.7%5.8 s136Not recordedRefund: YesInvoice: Yes↗
33SvenX

SvenX: Public-data score 67.5, with 85% weighted field coverage. Reported uptime: 94.9%; reported latency: 4.1 s.

Provider profile →
67.5Coverage 85%94.9%4.1 s19Not recordedRefund: YesInvoice: Yes↗
34aiaimark

aiaimark: Public-data score 45.2, with 100% weighted field coverage. Reported uptime: 69.3%; reported latency: 4 s.

Provider profile →
45.2Coverage 100%69.3%4 s425 / 5 (2)Refund: YesInvoice: No↗
35sidrune

sidrune: Public-data score 45.6, with 85% weighted field coverage. Reported uptime: 82.1%; reported latency: 7.4 s.

Provider profile →
45.6Coverage 85%82.1%7.4 s59Not recordedRefund: YesInvoice: Yes↗
36jcode

jcode: Public-data score 56.5, with 85% weighted field coverage. Reported uptime: 93.1%; reported latency: 5.9 s.

Provider profile →
56.5Coverage 85%93.1%5.9 s26Not recordedRefund: YesInvoice: No↗
37灵途

灵途: Public-data score 64.6, with 80% weighted field coverage. Reported uptime: 95.5%; reported latency: 6.1 s.

Provider profile →
64.6Coverage 80%95.5%6.1 s41Not recordedRefund: Not recordedInvoice: Yes↗
38FlyRouter

FlyRouter: Public-data score 73.7, with 85% weighted field coverage. Reported uptime: 99.5%; reported latency: 4.4 s.

Provider profile →
73.7Coverage 85%99.5%4.4 s47Not recordedRefund: YesInvoice: Yes↗
39onepay

onepay: Public-data score 62.8, with 80% weighted field coverage. Reported uptime: 94.3%; reported latency: 6 s.

Provider profile →
62.8Coverage 80%94.3%6 s43Not recordedRefund: Not recordedInvoice: Yes↗
40SaTuoAPI

SaTuoAPI: Public-data score 63.8, with 100% weighted field coverage. Reported uptime: 96.4%; reported latency: 6.7 s.

Provider profile →
63.8Coverage 100%96.4%6.7 s395 / 5 (1)Refund: NoInvoice: Yes↗

Model index

Start with the model you need

These are keyword-based research shortlists, not verified compatibility lists. Check the exact model ID, protocol and account limits before paying.

GPT

Check the advertised GPT route, not just its label.

279 matching recordsExplore model →

Claude

Check Claude-specific behavior in your intended integration.

253 matching recordsExplore model →

Codex

Clarify what the provider means by Codex access.

92 matching recordsExplore model →

Gemini

Validate the Gemini request types you actually need.

93 matching recordsExplore model →

DeepSeek

Distinguish DeepSeek model claims from deployment details.

104 matching recordsExplore model →

GLM

Check the GLM version and deployment arrangement.

78 matching recordsExplore model →

Qwen

Identify the specific Qwen variant behind the listing.

13 matching recordsExplore model →

Kimi

Verify Kimi claims against complete requests.

74 matching recordsExplore model →

AI Ranking

Choose evidence before an API provider

AI Ranking is a public-data index, not a laboratory benchmark. Use its listings to build a shortlist, identify missing information, and prepare your own checks. A high position does not establish model quality, security, or suitability for your workload.

Start with your actual use case

A small prototype

Prioritize a low spending limit, understandable documentation, and easy key revocation. Test one representative request before funding an account.

An interactive product

Check streaming, peak-hour errors, and time to the first useful response. A published latency figure cannot describe the whole user experience.

A production workflow

Require written data-handling terms, operational contacts, and a tested fallback. Treat purchasing, privacy, and technical acceptance as separate decisions.

Guides

Practical buying guides

Turn directory signals into questions, controlled trials, and explicit acceptance criteria.

Questions before you shortlist

Does AI Ranking test every provider?

No. The index organizes public records; its score is not an independent performance benchmark or a security assessment.

Does a higher score mean better answers?

No. The score combines listing signals, not task accuracy. Evaluate output quality with examples and acceptance criteria relevant to your application.

Are uptime and latency directly comparable?

Not necessarily. Records are snapshots with an unknown measurement window; regions, request types, and collection methods may differ.

Does a topic listing confirm model support?

No. Topics use keyword matches, not capability verification. Confirm exact models, versions, features, and access restrictions with the operator.

Is missing information a failed check?

No. Missing signals are omitted from the weighted mean and reduce coverage. They remain questions to investigate, not proof of failure.

Can I rely on a refund or invoice badge?

Only as a lead. Obtain the current written terms, applicable legal entity, eligibility conditions, and any deadlines before paying.