Hermes Agent
Hermes Agent is an open-source, self-improving AI agent by Nous Research that runs persistently with memory across sessions, and builds reusable skills from experience. It comes with 40+ built-in tools, including web search, browser automation, and vision, plus scheduled automations and subagents.
https://hermes-agent.nousresearch.com/ →
4.65T
Tokens (30d)
84.7M
Requests
$7.07M
Monthly cost
$1.73
Blended $/M tokens
Primary driver
Xiaomi: MiMo-V2-Pro is Hermes Agent's top model at 1.69T — 36.3% of total token volume and 37.2% of monthly cost ($2.63M).
Top-20 model breakdown
Ranked by monthly cost. Prices from OpenRouter's live catalog. Blended $/M = price_in × 72% + price_out × 28%.
| # | Model | Vendor | Tokens | % of vol. | $/M in | $/M out | Blended $/M | Monthly cost | % of spend |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Xiaomi: MiMo-V2-Pro xiaomi/mimo-v2-pro-20260318 | xiaomi | 1.69T | 36.3% | $1.00 | $3.00 | $1.56 | $2.63M | 37.2% |
| 2 | Anthropic: Claude Opus 4.6 anthropic/claude-4.6-opus-20260205 | anthropic | 124.1B | 2.7% | $5.00 | $25.00 | $10.60 | $1.32M | 18.6% |
| 3 | Anthropic: Claude Opus 4.7 anthropic/claude-4.7-opus-20260416 | anthropic | 55.8B | 1.2% | $5.00 | $25.00 | $10.60 | $591K | 8.4% |
| 4 | Anthropic: Claude Sonnet 4.6 anthropic/claude-4.6-sonnet-20260217 | anthropic | 91.3B | 2.0% | $3.00 | $15.00 | $6.36 | $581K | 8.2% |
| 5 | OpenAI: GPT-5.4 openai/gpt-5.4-20260305 | openai | 83.4B | 1.8% | $2.50 | $15.00 | $6.00 | $500K | 7.1% |
| 6 | Qwen: Qwen3.6 Plus qwen/qwen3.6-plus-04-02 | qwen | 610.3B | 13.1% | $0.33 | $1.95 | $0.78 | $476K | 6.7% |
| 7 | MiniMax: MiniMax M2.7 minimax/minimax-m2.7-20260318 | minimax | 395.0B | 8.5% | $0.30 | $1.20 | $0.55 | $218K | 3.1% |
| 8 | Z.ai: GLM 5.1 z-ai/glm-5.1-20260406 | z-ai | 103.8B | 2.2% | $1.05 | $3.50 | $1.74 | $180K | 2.5% |
| 9 | MoonshotAI: Kimi K2.6 moonshotai/kimi-k2.6-20260420 | moonshotai | 81.7B | 1.8% | $0.74 | $4.66 | $1.84 | $150K | 2.1% |
| 10 | Google: Gemini 3 Flash Preview google/gemini-3-flash-preview-20251217 | 102.2B | 2.2% | $0.50 | $3.00 | $1.20 | $123K | 1.7% | |
| 11 | OpenAI: GPT-5.4 Mini openai/gpt-5.4-mini-20260317 | openai | 38.7B | 0.8% | $0.75 | $4.50 | $1.80 | $70K | 1.0% |
| 12 | NVIDIA: Nemotron 3 Super nvidia/nemotron-3-super-120b-a12b-20230311 | nvidia | 231.0B | 5.0% | $0.09 | $0.45 | $0.19 | $44K | 0.6% |
| 13 | MoonshotAI: Kimi K2.5 moonshotai/kimi-k2.5-0127 | moonshotai | 50.1B | 1.1% | $0.44 | $2.00 | $0.88 | $44K | 0.6% |
| 14 | Qwen: Qwen3.6 Plus qwen/qwen3.6-plus-preview | qwen | 51.9B | 1.1% | $0.33 | $1.95 | $0.78 | $40K | 0.6% |
| 15 | StepFun: Step 3.5 Flash stepfun/step-3.5-flash | stepfun | 198.9B | 4.3% | $0.10 | $0.30 | $0.16 | $31K | 0.4% |
| 16 | MiniMax: MiniMax M2.5 minimax/minimax-m2.5-20260211 | minimax | 71.0B | 1.5% | $0.15 | $1.15 | $0.43 | $31K | 0.4% |
| 17 | Google: Gemini 2.5 Flash google/gemini-2.5-flash | 32.0B | 0.7% | $0.30 | $2.50 | $0.92 | $29K | 0.4% | |
| 18 | DeepSeek: DeepSeek V3.2 deepseek/deepseek-v3.2-20251201 | deepseek | 50.3B | 1.1% | $0.25 | $0.38 | $0.29 | $14K | 0.2% |
| 19 | Arcee AI: Trinity Large Preview arcee-ai/trinity-large-preview | arcee-ai | 28.9B | 0.6% | $0.15 | $0.45 | $0.23 | $7K | 0.1% |
| 20 | openrouter/elephant-alpha openrouter/elephant-alpha | openrouter | 73.8B | 1.6% | — | — | — | — | 0.0% |
Cross-references
- Full leaderboard: OpenRouter app leaderboard
- Inverted view: Which models AI agents actually use
- Related benchmarks: /agentic
- Source: openrouter.ai/apps/hermes-agent
Know Hermes Agent better than we do?
If these numbers don't match what you see from inside this app, tell us. We reply within 48 hours and update the analysis.