New agent

Identity

Lane assignment

Optional now — you can also create the agent empty and patch models afterwards from its page.

• Fast lane

  Cheap, high-volume classification, extraction, and short tool loops.

  Weight
  Leave emptyMistral: Mistral Nemo — $0.029/MTokinclusionAI: Ling 3.0 Flash VL — $0.021/MTokinclusionAI: Ling 3.0 Flash — $0.021/MTokSao10K: Llama 3 8B Lunaris — $0.04/MTokOpenAI: gpt-oss-20b — $0.018/MTokNex AGI: Nex-N2.5-Mini — $0.025/MTokIBM: Granite 4.0 Micro — $0.017/MTokMistral: Mistral Small 3 — $0.05/MTokMeta: Llama 3.1 8B Instruct — $0.05/MTokOpenAI: gpt-oss-20b (batch) — $0.024/MTokMistral: Ministral 3 8B 2512 (batch) — $0.075/MTokGoogle: Gemma 3 4B — $0.05/MTokQwen: Qwen3.7 Flash — $0.03/MTokinclusionAI: Ling 3.0 Flash Sante — $0.042/MTokinclusionAI: Ling 3.0 Flash Fin — $0.042/MTokOpenAI: gpt-oss-120b (batch) — $0.0296/MTokDeepSeek: DeepSeek Flash Latest — $0.0348/MTokAmazon: Nova Micro 1.0 — $0.035/MTokInference.net: Schematron V2 Turbo — $0.03/MTokPoolside: Laguna XS 2.1 — $0.06/MTokCohere: Command R7B (12-2024) — $0.0375/MTokInception: Mercury 2.5 — $0.04/MTokMythoMax 13B — $0.08/MTokReka Edge — $0.1/MTokMistral: Ministral 3 3B 2512 — $0.1/MTokGoogle: Gemma 3 12B — $0.05/MTokOpenAI: gpt-oss-120b — $0.037/MTokMicrosoft: Phi 4 — $0.07/MTokTencent: Hy-MT2-1.8B — $0.044/MTokOpenAI: GPT-5 Nano (batch) — $0.025/MTokMeta: Llama 3.2 1B Instruct — $0.027/MTokUpstage: Solar Mini 4 — $0.05/MTokQwen: Qwen3.5-9B — $0.1/MTokGoogle: Gemini 2.5 Flash Lite (batch) — $0.05/MTokOpenAI: GPT-4.1 Nano (batch) — $0.05/MTokZ.ai: GLM 5.3 Flash (batch) — $0.06/MTokNVIDIA: Nemotron 3.5 Lightning — $0.07/MTokPoolside: Laguna S 2.1 — $0.09/MTokInference.net: Schematron V2 Small — $0.05/MTokGoogle: Gemma 4 26B A4B — $0.0675/MTokAnthropic: Claude Haiku 5.5 (batch) — $0.05/MTokOpenAI: GPT-6 Luna Pro (batch) — $0.05/MTokOpenAI: GPT-6 Luna (batch) — $0.05/MTokNVIDIA: Nemotron 3 Nano 30B A3B — $0.06/MTokMistral: Ministral 3 8B 2512 — $0.15/MTokAmazon: Nova Lite 1.0 — $0.06/MTokMeta: Muse Spark 1.2 Contributor — $0.1/MTokMeta: Muse Spark 1.3 Contributor — $0.1/MTokByteDance: UI-TARS 7B — $0.1/MTokReka Flash 3 — $0.1/MTokQwen: Qwen2.5 7B Instruct — $0.1/MTokIBM: Granite 4.2 8B — $0.06/MTokNex AGI: Nex-N2.5-Pro — $0.075/MTokQwen: Qwen3.5-Flash — $0.065/MTokMistral: Mistral Small 3.2 24B — $0.09375/MTokQwen: Qwen3 Coder 30B A3B Instruct — $0.07/MTokMeta: Llama Guard 4 12B — $0.18/MTokQwen: Qwen3 14B — $0.12/MTokQwen: Qwen3 32B — $0.08/MTokTencent: Hy-MT2-30B-A3B — $0.074/MTok

• Balanced lane

  Default lane for everyday reasoning and multi-step agent work.

  Weight
  Leave emptyMistral: Mistral Nemo — $0.029/MTokinclusionAI: Ling 3.0 Flash VL — $0.021/MTokinclusionAI: Ling 3.0 Flash — $0.021/MTokSao10K: Llama 3 8B Lunaris — $0.04/MTokOpenAI: gpt-oss-20b — $0.018/MTokNex AGI: Nex-N2.5-Mini — $0.025/MTokIBM: Granite 4.0 Micro — $0.017/MTokMistral: Mistral Small 3 — $0.05/MTokMeta: Llama 3.1 8B Instruct — $0.05/MTokOpenAI: gpt-oss-20b (batch) — $0.024/MTokMistral: Ministral 3 8B 2512 (batch) — $0.075/MTokGoogle: Gemma 3 4B — $0.05/MTokQwen: Qwen3.7 Flash — $0.03/MTokinclusionAI: Ling 3.0 Flash Sante — $0.042/MTokinclusionAI: Ling 3.0 Flash Fin — $0.042/MTokOpenAI: gpt-oss-120b (batch) — $0.0296/MTokDeepSeek: DeepSeek Flash Latest — $0.0348/MTokAmazon: Nova Micro 1.0 — $0.035/MTokInference.net: Schematron V2 Turbo — $0.03/MTokPoolside: Laguna XS 2.1 — $0.06/MTokCohere: Command R7B (12-2024) — $0.0375/MTokInception: Mercury 2.5 — $0.04/MTokMythoMax 13B — $0.08/MTokReka Edge — $0.1/MTokMistral: Ministral 3 3B 2512 — $0.1/MTokGoogle: Gemma 3 12B — $0.05/MTokOpenAI: gpt-oss-120b — $0.037/MTokMicrosoft: Phi 4 — $0.07/MTokTencent: Hy-MT2-1.8B — $0.044/MTokOpenAI: GPT-5 Nano (batch) — $0.025/MTokMeta: Llama 3.2 1B Instruct — $0.027/MTokUpstage: Solar Mini 4 — $0.05/MTokQwen: Qwen3.5-9B — $0.1/MTokGoogle: Gemini 2.5 Flash Lite (batch) — $0.05/MTokOpenAI: GPT-4.1 Nano (batch) — $0.05/MTokZ.ai: GLM 5.3 Flash (batch) — $0.06/MTokNVIDIA: Nemotron 3.5 Lightning — $0.07/MTokPoolside: Laguna S 2.1 — $0.09/MTokInference.net: Schematron V2 Small — $0.05/MTokGoogle: Gemma 4 26B A4B — $0.0675/MTokAnthropic: Claude Haiku 5.5 (batch) — $0.05/MTokOpenAI: GPT-6 Luna Pro (batch) — $0.05/MTokOpenAI: GPT-6 Luna (batch) — $0.05/MTokNVIDIA: Nemotron 3 Nano 30B A3B — $0.06/MTokMistral: Ministral 3 8B 2512 — $0.15/MTokAmazon: Nova Lite 1.0 — $0.06/MTokMeta: Muse Spark 1.2 Contributor — $0.1/MTokMeta: Muse Spark 1.3 Contributor — $0.1/MTokByteDance: UI-TARS 7B — $0.1/MTokReka Flash 3 — $0.1/MTokQwen: Qwen2.5 7B Instruct — $0.1/MTokIBM: Granite 4.2 8B — $0.06/MTokNex AGI: Nex-N2.5-Pro — $0.075/MTokQwen: Qwen3.5-Flash — $0.065/MTokMistral: Mistral Small 3.2 24B — $0.09375/MTokQwen: Qwen3 Coder 30B A3B Instruct — $0.07/MTokMeta: Llama Guard 4 12B — $0.18/MTokQwen: Qwen3 14B — $0.12/MTokQwen: Qwen3 32B — $0.08/MTokTencent: Hy-MT2-30B-A3B — $0.074/MTok

• Deep lane

  Expensive frontier models for planning, long context, and hard reasoning.

  Weight
  Leave emptyMistral: Mistral Nemo — $0.029/MTokinclusionAI: Ling 3.0 Flash VL — $0.021/MTokinclusionAI: Ling 3.0 Flash — $0.021/MTokSao10K: Llama 3 8B Lunaris — $0.04/MTokOpenAI: gpt-oss-20b — $0.018/MTokNex AGI: Nex-N2.5-Mini — $0.025/MTokIBM: Granite 4.0 Micro — $0.017/MTokMistral: Mistral Small 3 — $0.05/MTokMeta: Llama 3.1 8B Instruct — $0.05/MTokOpenAI: gpt-oss-20b (batch) — $0.024/MTokMistral: Ministral 3 8B 2512 (batch) — $0.075/MTokGoogle: Gemma 3 4B — $0.05/MTokQwen: Qwen3.7 Flash — $0.03/MTokinclusionAI: Ling 3.0 Flash Sante — $0.042/MTokinclusionAI: Ling 3.0 Flash Fin — $0.042/MTokOpenAI: gpt-oss-120b (batch) — $0.0296/MTokDeepSeek: DeepSeek Flash Latest — $0.0348/MTokAmazon: Nova Micro 1.0 — $0.035/MTokInference.net: Schematron V2 Turbo — $0.03/MTokPoolside: Laguna XS 2.1 — $0.06/MTokCohere: Command R7B (12-2024) — $0.0375/MTokInception: Mercury 2.5 — $0.04/MTokMythoMax 13B — $0.08/MTokReka Edge — $0.1/MTokMistral: Ministral 3 3B 2512 — $0.1/MTokGoogle: Gemma 3 12B — $0.05/MTokOpenAI: gpt-oss-120b — $0.037/MTokMicrosoft: Phi 4 — $0.07/MTokTencent: Hy-MT2-1.8B — $0.044/MTokOpenAI: GPT-5 Nano (batch) — $0.025/MTokMeta: Llama 3.2 1B Instruct — $0.027/MTokUpstage: Solar Mini 4 — $0.05/MTokQwen: Qwen3.5-9B — $0.1/MTokGoogle: Gemini 2.5 Flash Lite (batch) — $0.05/MTokOpenAI: GPT-4.1 Nano (batch) — $0.05/MTokZ.ai: GLM 5.3 Flash (batch) — $0.06/MTokNVIDIA: Nemotron 3.5 Lightning — $0.07/MTokPoolside: Laguna S 2.1 — $0.09/MTokInference.net: Schematron V2 Small — $0.05/MTokGoogle: Gemma 4 26B A4B — $0.0675/MTokAnthropic: Claude Haiku 5.5 (batch) — $0.05/MTokOpenAI: GPT-6 Luna Pro (batch) — $0.05/MTokOpenAI: GPT-6 Luna (batch) — $0.05/MTokNVIDIA: Nemotron 3 Nano 30B A3B — $0.06/MTokMistral: Ministral 3 8B 2512 — $0.15/MTokAmazon: Nova Lite 1.0 — $0.06/MTokMeta: Muse Spark 1.2 Contributor — $0.1/MTokMeta: Muse Spark 1.3 Contributor — $0.1/MTokByteDance: UI-TARS 7B — $0.1/MTokReka Flash 3 — $0.1/MTokQwen: Qwen2.5 7B Instruct — $0.1/MTokIBM: Granite 4.2 8B — $0.06/MTokNex AGI: Nex-N2.5-Pro — $0.075/MTokQwen: Qwen3.5-Flash — $0.065/MTokMistral: Mistral Small 3.2 24B — $0.09375/MTokQwen: Qwen3 Coder 30B A3B Instruct — $0.07/MTokMeta: Llama Guard 4 12B — $0.18/MTokQwen: Qwen3 14B — $0.12/MTokQwen: Qwen3 32B — $0.08/MTokTencent: Hy-MT2-30B-A3B — $0.074/MTok

Weighted request counts are derived as weight × 2000 requests/day for the projection.