Skip to main content
All models are private. Zero data retention — no prompts, inputs, or outputs are stored or logged.

Chat Models

ModelIDCostContextAdded
GPT-5gpt-5$0.006/req128KJan 2026
GPT-5 Minigpt-5-mini$0.003/req128KJan 2026
Claude Opus 4.6claude-opus-4.6$0.030/req200KMar 2026
Claude Sonnet 4.6claude-sonnet-4.6$0.015/req200KMar 2026
Claude Sonnet 4.5claude-sonnet-4.5$0.006/req200KDec 2025
Claude Haiku 4.5claude-haiku-4.5$0.006/req200KDec 2025
o3-minio3-mini$0.006/req128KFeb 2026
Gemini 2.5 Progemini-2.5-pro$0.006/req1MFeb 2026
Gemini 2.5 Flashgemini-2.5-flash$0.003/req1MFeb 2026
Gemini 3 Progemini-3-pro$0.006/req1MMar 2026
Gemini 3.1 Progemini-3.1-pro$0.006/req1MMar 2026
Gemini 3 Flashgemini-3-flash$0.003/req1MMar 2026
Grok 4grok-4$0.006/req128KMar 2026
DeepSeek V3deepseek-v3$0.003/req128KNov 2025
Kimi K2kimi-k2$0.006/req128KFeb 2026
Kimi K2.5kimi-k2.5$0.006/req128KMar 2026
Mistral Largemistral-large$0.006/req128KJan 2026
Llama 4 Maverickllama-4-maverick$0.006/req128KMar 2026
Llama 4 Scoutllama-4-scout$0.003/req128KMar 2026
QwQ 32Bqwq-32b$0.003/req32KJan 2026
GLM 5glm-5$0.003/req128KFeb 2026
MiniMax M2.5minimax-m2.5$0.003/req128KJan 2026
Ninja 1ninja-1$0.003/req128KOct 2025
Uncensored AIuncensored-ai$0.003/req128KOct 2025

Image Models

ModelIDCostSpeedAdded
FLUX Kontext Maxflux-kontext-max$0.10/img~5-15sMar 2026
FLUX.2 Flexflux-2-flex$0.08/img~5-10sMar 2026
FLUX.1 Pro Ultraflux-1-pro-ultra$0.08/img~8-15sDec 2025
Recraft V3recraft-v3$0.08/img~5sNov 2025
Google Imagen 4google-imagen-4$0.08/img~15sMar 2026
Nano Banana Pronano-banana-pro$0.08/img~5sFeb 2026
Seedreamseedream$0.08/img~8sMar 2026
FLUX.2 Proflux-2-pro$0.05/img~3-5sFeb 2026
FLUX Kontext Proflux-kontext-pro$0.05/img~5sMar 2026
Nano Banana 2nano-banana-2$0.05/img~3sJan 2026
FLUX.1 Fillflux-1-fill$0.05/img~5sDec 2025
FLUX.2 Kleinflux-2-klein$0.03/imgunder 2sFeb 2026
Nano Banananano-banana$0.03/img~2sOct 2025

Video Models

ModelIDCostDurationSpeedAdded
Runway Gen-4.5runway-gen4.5$5.00/vid5-10s~1-3 minMar 2026
Veo 3.1veo-3.1$5.00/vid4-8s~3-5 minMar 2026
Veo 3.1 Fastveo-3.1-fast$3.00/vid4-8s~1-2 minMar 2026
Google Veo 2google-veo-2$3.00/vid5-8s~40sJan 2026
Veo 3 Fastgoogle-veo-3-fast$3.00/vid4-8s~1-2 minMar 2026
Seedance 2seedance-2$3.00/vid5-15s~1-2 minFeb 2026
Kling Videokling-video$3.00/vid5-10s~3-4 minFeb 2026
Runway Gen-4 Turborunway-gen4-turbo$3.00/vid5-10s~30-60sMar 2026

Smart Routing

IDStrategyBilled at
autoBest model per task typeResolved model’s rate
auto-fastLowest latencyResolved model’s rate
auto-cheapLowest costResolved model’s rate
auto-qualityHighest qualityResolved model’s rate
ensemble3-model consensus$0.040/req
ensemble-qualityPremium consensus$0.050/req
How smart routing works