Product Models Pricing
model catalog

Leading open-source models

wylon Token Factory serves the leading open-source model families — MiniMax, Kimi, GLM, Qwen, DeepSeek — all running on wylon's GPU super-node architecture behind a single unified API.

Frontiers

MiniMax

MiniMax

Long-context, productivity-grade workloads — document processing, summarization, multi-turn business conversations.

FlagshipMiniMax-M2.7 / M2.5
ContextLong-form / structured
Use casesEnterprise knowledge / long documents
Moonshot

Kimi

Solid on multimodal, ultra-long context, and code — a popular foundation for agents and developer tooling.

FlagshipKimi-K2.6 / K2.5
ContextLong dialogue / repos
Use casesAgent / IDE / research
Zhipu · Z.ai

GLM

Strong on Chinese corpora — covers general chat, tool use, and agentic workflows.

FlagshipGLM-5.1
ContextGeneral chat / tool calls
Use casesCustomer support / Chinese-language workloads
Qwen

Qwen

Full lineup from on-device tiny models to flagship MoEs, balancing performance and cost.

FlagshipQwen3.6
ContextGeneral / multilingual / multimodal
Use casesContent / RAG / vision
DeepSeek

DeepSeek

Reputation for reasoning, code, and math — its MoE line offers excellent price-performance.

FlagshipDeepSeek-V4
ContextReasoning / code / chain-of-thought
Use casesHigh-throughput batch / agent planning

And more

See the full list, context lengths, and pricing in the model matrix below and on the pricing page.

View pricing →

Other

Refer to the live service for the authoritative model list. See the pricing page or your console for current pricing.

ModelsFamilyContextBest for
Kimi-K2.5
General chat · long context
Kimi
128k
Long documents, knowledge Q&A, multi-turn chat
Kimi-K2.6
General chat · enhanced
Kimi
128k
Complex reasoning, creative writing, expert Q&A
MiniMax-M2.5
General chat · cost-effective
MiniMax
256k
Bulk summarization, customer support, content
MiniMax-M2.7
General chat · enhanced
MiniMax
256k
High-quality generation, multi-turn chat, tool use
GLM-5.1
Chinese general · tool use
GLM
128k
Code generation, structured output, customer support
Qwen3.6-35B-A3B
Flagship MoE · reasoning
Qwen
256k
Data analysis, code, workflow orchestration
Qwen3.6-27B
Dense · reasoning
Qwen
256k
Complex reasoning, RAG, IDE assistants
DeepSeek-V4-Flash
Fast · low cost
DeepSeek
64k
High-concurrency inference, summarization, classification
Access method

Single wylon API

All models share the same endpoint — just switch the model field to compare families.

Get started → Switch to wylon → Pricing →

FAQ

Do I need to change my code when switching models?

Just swap the model field — the request shape stays the same. If you rely on tool calls or structured output, we recommend running a regression pass after switching.

Will I be notified when new models go live?

Yes. wylon posts release announcements in the dashboard, and major changes are sent ahead of time.

Can enterprises request private access?

Please contact us — our solutions team can help scope a plan with you.