Jurisdiction-enforced inferenceAccounts are reviewed per jurisdiction and billed from the first token. No free credits, no promotional tier.Contact us for a quotation

Catalogue

Every endpoint, with its serving region

A model is not one thing. The same weights on different infrastructure carry different legal positions, so each deployment gets its own row: context window, rate, modalities and the region that serves it. Every request to this catalogue is answered from one of these endpoints, or refused.

Filter the catalogue

Showing 44 of 44 entries

Models, rates and serving regions
ModelProviderContextInput, per 1MOutput, per 1MModalitiesRegionsResidency
Qwen3.8-Max2026-08-02Alibaba Qwen1M$2.00$6.00Text · Image
ap-southeast-1ap-east-1ap-southeast-5
Cannot guarantee
Qwen3.8-Flash2026-08-26Alibaba Qwen1M$0.15$0.47Text · Image · Video
ap-southeast-1ap-east-1ap-southeast-5
Cannot guarantee
Qwen3.7-Plus2026-06-01Alibaba Qwen1M0.40 (0-256K input); 1.20 (256K-1M input)1.60 (0-256K input); 4.80 (256K-1M input)Text · Image
ap-southeast-1ap-east-1ap-southeast-5
Cannot guarantee
ERNIE 5.1UnverifiedBaidu ERNIE128KUnverifiedUnverifiedText
cn-beijingap-east-1
In region
ERNIE 5.02025-11Baidu ERNIE128KUnverifiedUnverifiedText · Image · Audio · Video
cn-beijingap-east-1
In region
ERNIE 4.5 Turbo VL2025-08-07Baidu ERNIE128KUnverifiedUnverifiedText · Image
cn-beijingap-east-1
In region
Doubao Seed 2.1 Pro2026-09-15ByteDance (Doubao / Seed)1.02MUnverifiedUnverifiedText · Image · Video
cn-beijingap-southeast-1ap-east-1
In region
Doubao Seed 2.1 Lite2026-09-15ByteDance (Doubao / Seed)1.02MUnverifiedUnverifiedText · Image · Video
cn-beijingap-southeast-1ap-east-1
In region
Doubao Seed 2.1 Turbo2026-06-28ByteDance (Doubao / Seed)262KUnverifiedUnverifiedText · Image · Video
cn-beijingap-southeast-1ap-east-1
In region
DeepSeek V4 Pro2026-08-13DeepSeek1M$0.3$1.2Text
CNcn-beijingap-southeast-1
In region
DeepSeek V4.1 Flash2026-09-10DeepSeek1M$0.3$1.2Text · Image
CNcn-beijingap-southeast-1
In region
MiniMax-M32026-06-01MiniMax1M$0.3$1.2Text · Image · Video
ap-east-1undisclosed
Cannot guarantee
MiniMax-M2.72026-03-18MiniMax205K$0.3$1.2Text
ap-east-1undisclosed
Cannot guarantee
Kimi K32026-07-16Moonshot AI (Kimi)1.05M$3.00$15.00Text · Image · Video
cn-beijingcn-hongkongap-southeast-1
In region
Kimi K2.62026-04-20Moonshot AI (Kimi)262K$0.95$4.00Text · Image · Video
cn-beijingcn-hongkongap-southeast-1
In region
Kimi K2.7 Code2026-06-16Moonshot AI (Kimi)262K$0.95$4.00Text · Image · Video
cn-beijingcn-hongkongap-southeast-1
In region
Step 5 Preview2026-09-20StepFun1MUnverifiedUnverifiedText · Image · Video
ap-east-1undisclosed
Cannot guarantee
Step 3.7 Flash2026-05-29StepFun256K$0.2$1.15Text · Image · Video
ap-east-1undisclosed
Cannot guarantee
Step 3.5 Flash2026-02-12StepFun256K$0.1$0.3Text
ap-east-1undisclosed
Cannot guarantee
Hy4 preview2026-08-28Tencent Hunyuan1.05MUnverifiedUnverifiedText
ap-guangzhouap-singaporeap-east-1
In region
Hy32026-07-06Tencent Hunyuan262KUnverifiedUnverifiedText
ap-guangzhouap-singaporeap-east-1
In region
GLM-5.3-Flash2026-08-26Z.ai (GLM)1M$0.15$0.5Text · Image · Video
ap-east-1undisclosed
Cannot guarantee
GLM-5.32026-08-14Z.ai (GLM)1M$1.4$4.4Text
ap-east-1undisclosed
Cannot guarantee
GLM-5.22026-06-16Z.ai (GLM)1M$1.4$4.4Text
ap-east-1undisclosed
Cannot guarantee
DeepSeek V4 Pro on AWSUnverifiedDeepSeek on AWS1MUnverifiedUnverifiedText
ap-east-1ap-southeast-5
Cannot guarantee
DeepSeek V4 Pro on GCPUnverifiedDeepSeek on GCP1MUnverifiedUnverifiedText
ap-east-1ap-southeast-5
Cannot guarantee
Qwen3.8-Max on AWSUnverifiedQwen on AWS1MUnverifiedUnverifiedText · Image
ap-east-1ap-southeast-5
Cannot guarantee
Claude Opus 5.52026-09-22Anthropic1M$4.00$20.00Text · Image
us-east-1ap-east-1
In region
Claude Sonnet 5.52026-09-28Anthropic1M$2.00$10.00Text · Image
us-east-1ap-east-1
In region
Claude Haiku 4.52025-10-15Anthropic200K$1.00$5.00Text · Image
us-east-1ap-east-1
In region
Gemini 3.8 Flash2026-09-02Google (Gemini)1.05M$0.75$3.75Text · Image · Audio · Video
ap-east-1undisclosed
Cannot guarantee
Gemini 3.5 Flash2026-05-19Google (Gemini)1.05M$1.5$9.00Text · Image · Audio · Video
ap-east-1undisclosed
Cannot guarantee
Gemini 3.1 Pro Preview2026-02-19Google (Gemini)1.05M2.00 (<=200K prompt); 4.00 (>200K)12.00 (<=200K prompt); 18.00 (>200K)Text · Image · Audio · Video
ap-east-1undisclosed
Cannot guarantee
Gemini 2.5 Pro2025-06-17Google (Gemini)1.05M1.25 (<=200K prompt); 2.50 (>200K)10.00 (<=200K prompt); 15.00 (>200K)Text · Image · Audio · Video
ap-east-1undisclosed
Cannot guarantee
Llama 4 Maverick2025-04-05Meta (Llama)1M$0.25$1.00Text · Image
us-east-1ap-east-1
In region
Llama 4 Scout2025-04-05Meta (Llama)10MUnverifiedUnverifiedText · Image
us-east-1ap-east-1
In region
Mistral Medium 3.52026-04-28Mistral AI256K$1.5$7.5Text · Image
ap-east-1undisclosed
Cannot guarantee
Mistral Small 42026-03-16Mistral AI256K$0.15$0.6Text · Image
ap-east-1undisclosed
Cannot guarantee
Mistral Large 32025-12-02Mistral AI256K$0.5$1.5Text · Image
ap-east-1undisclosed
Cannot guarantee
GPT-6 Astra2026-09-03OpenAI1.05M$10.00$50.00Text · Image
ap-east-1undisclosed
Cannot guarantee
GPT-6.1 Sol2026-09-29OpenAI1.05M$2.00$10.00Text · Image
ap-east-1undisclosed
Cannot guarantee
GPT-6 Luna2026-09-22OpenAI1.05M$0.1$0.5Text · Image
ap-east-1undisclosed
Cannot guarantee
Grok 4.72026-09-21xAI (Grok)500K$2.00$6.00Text · Image
us-east-1ap-east-1
In region
Grok 4.3UnverifiedxAI (Grok)1M$1.25$2.5Text · Image
us-east-1us-west-2eu-west-1
In region


44
Model entries
18
Providers
20
With a verified in-region endpoint
19
Open-weight entries

Hostable inside your own boundary


01

How the catalogue is organised

Three catalogue tiers

Featured

Current China-based labs, prioritised in this catalogue.

24 entries

Cross-border deployments

The same weights served by different infrastructure, which is a different legal position.

3 entries

Included

Frontier and open-weight models from European and American labs.

17 entries

Fields marked Unverified are unpublished by the provider.

02

Reading an entry

Column definitions

  1. 01

    Serving regions

    The regions from which a model can be reached, taken from provider documentation or a cloud region inventory. A blank means no region was published; it does not say the model is unavailable.

  2. 02

    Residency: cannot guarantee

    One or more facts needed to enforce a boundary are unpublished, usually the region. These entries cannot take jurisdiction-bound traffic.

  3. 03

    Input and output rates

    USD per million tokens. If pricing varies by input length, both published bands appear. If a provider prices only in local currency, the USD field stays unverified rather than being converted.

  4. 04

    Regions in this catalogue

    13 distinct region codes are used by 18 providers. A code describes where compute runs; it does not settle the legal mechanism. Your filing obligations remain yours.

Next

API key access

Billing starts with the first token. There is no free tier, and the endpoint is OpenAI-compatible.