Model registry

Hosted open models, one unified API.

Browse production-hosted model routes that run on approved GPU capacity. Developers keep one key and one base URL while SCX manages routing, deployment posture, and credit settlement behind the scenes.

Models25
Live20
Providers14
Production fit
Showing 25 / 25
01DeepSeek
DE

scx/deepseek-r1-32b

Reasoning · Agent planning, tool use, structured reasoning

Context128K
TTFTBalanced
Price1.45 Credit / 1K output tokens
02DeepSeek
DE

scx/deepseek-r1-70b

Reasoning · Hard reasoning, math, multi-step analysis

Context128K
TTFTDeep
Price1.95 Credit / 1K output tokens
03DeepSeek
DE

scx/deepseek-v3

General chat · Copilots, workflows, customer support

Context128K
TTFTFast
Price0.82 Credit / 1K output tokens
04Qwen
QW

scx/qwen3-235b-a22b

Reasoning · Hard reasoning, math, multi-step analysis

Context128K
TTFTDeep
Price2.20 Credit / 1K output tokens
05Qwen
QW

scx/qwen3-32b

Reasoning · Fast assistants, routing fallback, live UX

Context128K
TTFTFast
Price0.78 Credit / 1K output tokens
06Qwen
QW

scx/qwen2.5-72b-instruct

General chat · Copilots, workflows, customer support

Context128K
TTFTFast
Price0.95 Credit / 1K output tokens
07Qwen
QW

scx/qwen2.5-coder-32b

Coding · Code generation, review, migration plans

Context128K
TTFTFast
Price0.82 Credit / 1K output tokens
08Qwen
QW

scx/qwen2.5-vl-72b

Vision · Document vision, screenshots, visual reasoning

Context128K
TTFTBalanced
Price1.05 Credit / 1K output tokens
09Meta
ME

scx/llama-3.3-70b

Multilingual · Global apps and multilingual assistant routes

Context128K
TTFTFast
Price0.88 Credit / 1K output tokens
10Meta
ME

scx/llama-3.1-8b-instruct

General chat · Low-cost fallback routes and high-volume traffic

Context128K
TTFTVery fast
Price0.22 Credit / 1K output tokens
11Mistral
MI

scx/mistral-large-2

Coding · Code generation, review, migration plans

Context128K
TTFTBalanced
Price1.10 Credit / 1K output tokens
12Mistral
MI

scx/mixtral-8x22b

General chat · Copilots, workflows, customer support

Context64K
TTFTFast
Price0.78 Credit / 1K output tokens
13Google
GO

scx/gemma-3-27b-it

Vision · Multimodal apps, image understanding, assistants

Context128K
TTFTFast
Price0.58 Credit / 1K output tokens
14Google
GO

scx/gemma-2-9b-it

General chat · Low-cost fallback routes and high-volume traffic

Context8K
TTFTVery fast
Price0.16 Credit / 1K output tokens
15Zhipu
ZH

scx/glm-4-32b

General chat · Copilots, workflows, customer support

Context128K
TTFTFast
Price0.74 Credit / 1K output tokens
1601.AI
01

scx/yi-large

Multilingual · Global apps and multilingual assistant routes

Context200K
TTFTBalanced
Price0.92 Credit / 1K output tokens
17MiniMax
MI

scx/minimax-text-01

Long context · Long documents, research packs, policy analysis

Context1M
TTFTBalanced
Price1.25 Credit / 1K output tokens
18InternLM
IN

scx/internlm2.5-20b-chat

General chat · Fast assistants, routing fallback, live UX

Context32K
TTFTVery fast
Price0.36 Credit / 1K output tokens
19OpenGVLab
OP

scx/internvl3-38b

Vision · Document vision, screenshots, visual reasoning

Context128K
TTFTBalanced
Price0.96 Credit / 1K output tokens
20BAAI
BA

scx/bge-m3

Embeddings · Semantic search, retrieval, knowledge bases

Context8K
TTFTInstant
Price0.12 Credit / 1K input tokens
21Jina AI
JI

scx/jina-embeddings-v3

Embeddings · Memory layers, RAG pipelines, long retrieval

Context8K
TTFTInstant
Price0.10 Credit / 1K input tokens
22Mistral
MI

scx/e5-mistral-7b-instruct

Embeddings · Memory layers, RAG pipelines, long retrieval

Context32K
TTFTVery fast
Price0.18 Credit / 1K input tokens
23BAAI
BA

scx/bge-reranker-v2-m3

Rerank · Search quality, retrieval reranking, answer grounding

Context8K
TTFTVery fast
Price0.18 Credit / 1K input tokens
24Whisper
WH

scx/whisper-large-v3

Audio · Speech transcription and meeting intelligence

Context30 min
TTFTBalanced
Price0.20 Credit / minute
25Stability
ST

scx/stable-diffusion-xl

Image · Image generation, creative workflows, drafts

ContextImage
TTFTBalanced
Price0.65 Credit / image

Do not see the model you need?

Request a new route or ask for reserved deployment capacity for a private workload.