21+ frontier models, one stop

Full coverage of general reasoning, long context, multimodal, OCR and embeddings — versions track official releases

General & reasoning

Chat · reasoning · code · agents — flagship & lightweight tiers within each family, Tune picks the optimal one

DeepSeekDeep reasoning
QwenCode generation
GLMAgent orchestration
KimiExtra-long context
ClaudeComplex tasks
GPT seriesRich ecosystem
GeminiMultimodal
MiniMaxGeneral chat

Embeddings & retrieval

Embedding · Reranker — built for enterprise knowledge bases and semantic search

bge-m3Multilingual vectors
bge-large-zhChinese vectors
bge-rerankerPrecision reranking
jinaMultilingual embeddings
OpenAI embeddingGeneral embeddings
Typical scenarios: enterprise RAG · semantic search · document dedup · Q&A recall

Vision & multimodal

OCR · vision-language (VL) · video understanding

Qwen-VLImage-text understanding
GLM-OCRDocument parsing
DeepSeek-OCRLayout restoration
ClaudeVisual reasoning
GPTOmni-modal
Typical scenarios: invoice OCR · image review · long-video understanding · product recognition

Full model catalog

Uniform supply at ~15% off official list price · one API Key for everything

Leading Chinese models

DeepSeekQwenKimiGLMMiniMaxDoubaoERNIEHunyuanSparkBaichuanYi

Leading global models

GPT seriesClaudeGeminiLlamaMistral
Chat · everyday Q&A Deep reasoning · complex tasks Long context / Agent · orchestration Multimodal / OCR · vision Embeddings · retrieval Code models · engineering
* Models are continuously onboarded; actual availability is subject to the console

The Token economy — stop agonizing over model choice

The industrial era ran on electricity (kWh); the intelligence era runs on compute, delivered as tokens. Too many models, too many keys — leave the selection to TuneRouter.

OpenRouter
300+
TuneRouter
200+
Alibaba Bailian
152
SiliconFlow
107
Baidu Qianfan
103
Model counts across major LLM aggregation platforms · Source: public industry statistics & platform websites, as of March 2026
Jan 2024 Nationwide daily model calls began explosive growth Mar 2026 Token-based settlement is now mainstream · models in full bloom

New versions onboard within 72 hours

New releases from DeepSeek, Qwen, Claude, GPT and other vendors are callable the day they launch — no scheduling, no re-integration

One Key, every model

~15% off list price · new models within 72h · automatic routing & tuning

Contact Our Team
mail@TuneRouter.com