Free & cheap models for Hermes Agent
Every OpenRouter model with tool calling and 64,000+ tokens of context (what Hermes needs), free first, then the cheapest paid. Price and specs only, not quality.
Updated Nov 3, 2026, 9:59 AM CT. Source: OpenRouter models API and the Hermes Agent page on OpenRouter. Not affiliated with OpenRouter or Nous Research.
Quick answers
- What does a model need to work with Hermes Agent?
- At least 64,000 tokens of context, per the Hermes Agent quickstart. Every model listed here has that much context and supports tool calling on OpenRouter.
- Are the free models really free?
- Yes, $0 for input and output on OpenRouter. They're rate-limited, and some have an end date, which we show when OpenRouter lists one.
- How often does this list update?
- We read OpenRouter's public model list at most once an hour. It covers price and specs only, not how well each model works.
Free models with tool calling (14)
$0 in and out, sorted by context size. Free models are rate-limited; end dates shown when listed.
| Model | Model ID | Context |
|---|---|---|
| Thinking Machines: Inklinglisted until 2026-11-04 | thinkingmachines/inkling:free | 1M |
| Thinking Machines: Inkling Smalllisted until 2026-11-04 | thinkingmachines/inkling-small:free | 1M |
| NVIDIA: Nemotron 3 Ultra | nvidia/nemotron-3-ultra-550b-a55b:free | 1M |
| NVIDIA: Nemotron 3.5 Lightning | nvidia/nemotron-3.5-lightning:free | 1M |
| NVIDIA: Nemotron 3.5 Super VL | nvidia/nemotron-3.5-super-vl:free | 1M |
| Dots Studio: Dots3-Note Previewlisted until 2026-12-31 | dots-studio/dots-3-note-preview:free | 512K |
| Google: Gemma 4 26B A4B | google/gemma-4-26b-a4b-it:free | 262K |
| Google: Gemma 4 31B | google/gemma-4-31b-it:free | 262K |
| Recursive: Flint | recursive/flint-v1.0:free | 262K |
| XingChen: Xing4.0 29B A4B | xingchen-agi/xing4.0-29b-a4b:free | 262K |
| Cohere: North Mini Code | cohere/north-mini-code:free | 256K |
| NVIDIA: Nemotron 3 Nano Omni | nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free | 256K |
| LiquidAI: LFM2.5-2.6B | liquid/lfm-2.5-2.6b:free | 66K |
| Free Models Router(router: picks a free model per request) | openrouter/free | 200K |
Cheapest paid models with tool calling
Sorted by input plus output price, US dollars per million tokens. Batch-only variants left out.
| Model | Model ID | Input / 1M | Output / 1M | Context |
|---|---|---|---|---|
| Mistral: Mistral Nemo | mistralai/mistral-nemo | $0.029 | $0.03 | 131K |
| inclusionAI: Ling 3.0 Flash | inclusionai/ling-3.0-flash | $0.021 | $0.063 | 262K |
| Z.ai: GLM 5.3 Flash | z-ai/glm-5.3-flash | $0.031 | $0.085 | 1M |
| Meta: Llama 3.1 8B Instruct | meta-llama/llama-3.1-8b-instruct | $0.05 | $0.08 | 131K |
| Qwen: Qwen3.7 Flash | qwen/qwen3.7-flash | $0.03 | $0.13 | 1M |
| OpenAI: gpt-oss-20b | openai/gpt-oss-20b | $0.03 | $0.13 | 131K |
| Amazon: Nova Micro 1.0 | amazon/nova-micro-v1 | $0.035 | $0.14 | 128K |
| Inception: Mercury 2.5 | inception/mercury-2.5 | $0.04 | $0.15 | 260K |
| Mistral: Ministral 3 3B 2512 | mistralai/ministral-3b-2512 | $0.1 | $0.1 | 131K |
| Google: Gemma 3 12B | google/gemma-3-12b-it | $0.05 | $0.15 | 131K |
| IBM: Granite 4.2 8B | ibm-granite/granite-4.2-8b | $0.06 | $0.15 | 131K |
| PrismML: Ternary Bonsai 2 9B | prism-ml/ternary-bonsai-2-9b | $0.08 | $0.13 | 262K |
| inclusionAI: Ling 3.0 Flash VL | inclusionai/ling-3.0-flash-vl | $0.06 | $0.18 | 131K |
| Upstage: Solar Mini 4 | upstage/solar-mini4 | $0.05 | $0.2 | 524K |
| NVIDIA: Nemotron 3.5 Lightning | nvidia/nemotron-3.5-lightning | $0.07 | $0.2 | 262K |
| DeepSeek: DeepSeek V4 Flash 0423 | deepseek/deepseek-v4-flash | $0.091 | $0.182 | 1M |
| Mistral: Mistral Small 3.2 24B | mistralai/mistral-small-3.2-24b-instruct | $0.075 | $0.2 | 256K |
| inclusionAI: Ling 3.1 Flash | inclusionai/ling-3.1-flash | $0.075 | $0.225 | 1M |
| Amazon: Nova Lite 1.0 | amazon/nova-lite-v1 | $0.06 | $0.24 | 300K |
| NVIDIA: Nemotron 3 Nano 30B A3B | nvidia/nemotron-3-nano-30b-a3b | $0.06 | $0.24 | 262K |
What Hermes Agent users run most on OpenRouter
OpenRouter's ranking by tokens sent from Hermes Agent this month.
| # | Model | Tokens this month | Input / 1M | Output / 1M |
|---|---|---|---|---|
| 1 | DeepSeek V4.1 Flash | 11.4T | $0.045 | $0.5 |
| 2 | DeepSeek V4.1 Flash 1017 | 6.27T | $0.15 | $0.6 |
| 3 | GLM 5.3 Flash | 3.88T | $0.031 | $0.085 |
| 4 | Step 5 Preview | 3.19T | $1.00 | $2.70 |
| 5 | Solar Mini 4 | 2.63T | $0.05 | $0.2 |
| 6 | DeepSeek V4 Flash 0731 | 2.39T | $0.059 | $0.3 |
| 7 | MiniMax M3.1 Flash | 2.14T | $0.075 | $0.3 |
| 8 | GPT-6 Luna | 1.77T | $0.1 | $0.5 |
| 9 | Nemotron 3 Ultra | 1.67T | $0.6 | $2.40 |
| 10 | DeepSeek V4 Flash 0423 | 1.49T | $0.091 | $0.182 |
Monthly cost estimate
Uses today's OpenRouter list prices. The numbers below are placeholders, so put in your own. An agent turn resends the system prompt and tool list, so input tokens per message are usually far higher than what you type. Prompt caching and provider discounts aren't counted.
How the list is made
We read OpenRouter's public model list at most hourly. Free means $0 input and output. Tools means OpenRouter lists “tools” as supported. A daily copy lets the alert email spot new, removed or cheaper models.