Catalog
Providers and models Zintus routes to
57 providers · 100 models · 18 free · 22 routable today, 35 planned (not yet routable).
100 of 100 models
GPT-OSS 120B
OpenAI open-weight · free on Groq/OR
Gemma 4 31B
Multimodal · 140 languages · free
Gemma 4 26B MoE
Efficient MoE · 3.8B active params
Llama 4 Scout
10M context · longest free ctx ever
Llama 4 Maverick
Multimodal · 1M context · free
Llama 3.3 70B
Most-used free model on OR
DeepSeek R1
Best free reasoning model
DeepSeek V3
Strong general · free
Qwen3 Coder 480B
Best free coder globally · Jun 2026
Qwen 2.5 7B
Lightweight free · classification
Mistral Small
EU option · free
Nemotron 3 Ultra 550B
1M ctx · frontier reasoning · free
Nemotron 3 Super 120B
Agentic · 1M ctx · free
GLM-4.5-Air
Chinese/multilingual · free
Poolside Laguna
Coding agents · free
Owl Alpha
Tool use · agentic · free
MiMo V2.5
Top OR coding ranking Jun 2026
Phi-4
Microsoft open · tiny but capable
Ministral 3B
Cheapest Mistral · equal in/out
Command R7B
Cheapest first-party RAG model
GPT-OSS 20B
Ultra-cheap OpenAI open model
Llama 3.1 8B Instant
Cheapest + fastest · default T0
Nemotron 3 Nano 30B
Edge · tiny MoE · fast
Gemini 1.5 Flash
Legacy · still competitive
Nemotron 3 Super 120B
Cheapest 1M ctx paid model
GPT-4.1 Nano
Cheapest OpenAI · 1M context
Gemini 2.5 Flash Lite
Cheapest Google paid
Llama 4 Scout
Cheapest long-context paid option
Mistral 7B
Original · ultra-cheap
Embed v3 English
Best English embeddings
GLM-4 Plus
Budget GLM · equal in/out pricing
Mistral Nemo
Budget EU option
DeepSeek V4 Flash
Best cheap coder · MIT · 1M ctx
Yi Lightning
Fast · cheap · 01.AI
Llama 4 Maverick
Strong general open model
Command R
Grounded generation · citations
QwQ 32B
Alibaba reasoning · strong benchmarks
Command Nightly
Latest Cohere research build
Llama 3.2 11B Vision
Vision + text · multimodal T0
SOLAR Pro
Upstage AI · Korean/English bilingual
Gemma 4 31B (paid)
Equal input/output pricing
DeepSeek V3
Strong general · MIT
Grok 4.1 Fast
2M ctx · cheapest long-context
Sonar Small
Web-grounded · live citations
Step 3.7 Flash
Fast · multimodal · coding
Jamba 1.6 Mini
Hybrid SSM-Transformer · AI21
Gemini 2.5 Flash
Fast capable Google
Codestral
Code-specialized Mistral
Sonar
Balanced search-grounded
MiniMax M3
1M ctx · top OR coding rankings
MiniMax M2.7
Prior MiniMax · stable pricing
Yi Large
01.AI · multilingual
Qwen3.6 27B
Multimodal · image + video + text
GPT-4.1 Mini
Best budget OpenAI
Qwen3.7 Plus
1M context Qwen
Mistral Medium 3
Solid EU mid-range
Nemotron 4 340B
Reward model · RLHF training data
Jamba 1.6 Large
Large hybrid SSM-Transformer
Qwen3 235B MoE
Flagship open-weight · top OR volume
Qwen3.6 Plus
Balanced Qwen · 1M context
GLM-5.1
Z.AI coding flagship
DeepSeek R1
MIT reasoning · matches o1 benchmarks
Llama 3.3 70B
Fastest 70B · 394 TPS on Groq
Llama 3.1 70B
Reliable workhorse
DeepSeek V3.1
US-hosted DeepSeek · stable latency
Qwen3.5 397B MoE
Large MoE Qwen · strong reasoning
Hermes 3 70B
Strong instruction · agentic
Claude Haiku 3.5
Older Haiku · still available
Qwen 2.5 72B
Solid mid-size Qwen
Kimi K2.7 Code
Frontier coding · Moonshot AI
Kimi K2.6
Coding · Moonshot AI
Claude Haiku 4.5
Claude quality at T1 price
Grok Build 0.1
Coding specialist
Sonar Reasoning
Reasoning + web grounding
o3-mini
Budget reasoning
o4-mini
Latest budget reasoning
Mixtral 8x22B
MoE classic · strong benchmark
Gemini 2.5 Pro
Best Google reasoning value
Gemini 1.5 Pro
Legacy 2M context option
Grok 4.3
Flagship · real-time data · X access
Grok 4.20 (standard)
xAI production standard
Grok 4.20 Reasoning
Chain-of-thought reasoning
Grok 4.20 Multi-Agent
2M context · multi-agent orchestration
Gemini 3.5 Flash
Beats 3.1 Pro on coding · May 2026
DeepSeek V4 Pro
Hard reasoning + code
GPT-4.1
OpenAI flagship · 1M context
o3
Strong reasoning · 80% cheaper than launch
Gemini 3.1 Pro
Google flagship
Mistral Large 2
EU flagship · GDPR-native
Rerank v3
Reranking for RAG pipelines · $2/1K search units (not per-token)
Sonar Reasoning Pro
Deep reasoning + deep research
GPT-4o
Previous gen · multimodal
Command R+
RAG flagship · full retrieval stack
Claude Sonnet 4.6
Managed ceiling · best balance
Grok 4 Heavy
Multi-agent · HLE 50.7% · BYOK only
Sonar Pro
Deep research · 200K ctx
Llama 3.1 405B
Largest open model available
Claude Opus 4.8
Best coder · frontier · BYOK only
Claude Opus 4.7
Vision + long-horizon agents
GPT-5.5
Latest OpenAI frontier