AgMoDB
ModelsAgentsEvalsCompositesVisualizeIndustry
AgMoDB by @mistakeknot

Model picks

Current defaults by use case.

Reviewed Jul 24, 2026 · Benchmark, price, speed, and human preference.

Product

Production assistants and internal tools.

Default
Claude Sonnet 5 (Adaptive Reasoning, Max Effort)

Anthropic

AgMoBench 53.3$4.00/M81 tok/sEarly evidence

Current Sonnet balance of agency, quality, and cost.

View model
Value
Gemini 3.5 Flash-Lite

Google

AgMoBench 49.8$0.850/M327 tok/sEarly evidence

Fast, low-cost agentic throughput.

View model
Ceiling
Claude Opus 5 (Adaptive Reasoning, Max Effort)

Anthropic

AgMoBench 68.5$10.00/M53 tok/sEarly evidence

Highest-confidence production work.

View model
Browse all modelsCompare picks

Human frontier

See all
1Claude Opus 5 (Adaptive Reasoning, Max Effort)AnthropicHuman Frontier 96.2$10.00/M53 tok/s2Kimi K3KimiHuman Frontier 95.8$6.00/M39 tok/s3Anthropic: Claude Opus 4.7AnthropicHuman Frontier 95.5$10.00/M—4Claude Opus 4.6 (Non-reasoning, High Effort)AnthropicHuman Frontier 95.2$10.00/M—5GLM-5.2 (max)Z AIHuman Frontier 94.7$2.15/M119 tok/s6Qwen3.7 MaxAlibabaHuman Frontier 94.2$3.75/M—

Worth discovering

Frontier value

Kimi K3

Kimi

New open-frontier pressure.

Cheap reasoning

GLM-5.2 (max)

Z AI

Recent reasoning price/performance.

Fast batch work

Gemini 3.5 Flash-Lite

Google

Fast, cheap high-throughput lane.

Open pressure

Qwen3.7 Plus

Alibaba

Open-ish frontier compression.