log inget an API key

Hermes Agent with SayGM

Nous Research's self-improving agent, running on SayGM's models.

Model

Claude Opus 5.5Best quality

In $4.00$3.50Out $20.00$17.51/1M

Sign in to fill these snippets with a SayGM key for Hermes Agent.

Sign in

Set up in one command

Signs you in, creates a key for this tool, installs the tool if it is missing, and writes its config. Run npx saygm uninstall hermes to put the files back.

  1. 1Run the installer
    bash
    npx saygm install hermes --model claude-opus-5-5

Requires Node.js 22 or later.

SayGM CLI on npm

Setup for Claude models

Use this setup with the models listed under Claude below.

  1. 1~/.hermes/config.yaml
    yaml
    providers:
  saygm-claude:
    base_url: https://api.saygm.com
    key_env: SAYGM_API_KEY
    api_mode: anthropic_messages
    models:
      - claude-opus-5-5
      - claude-sonnet-5

model:
  provider: custom:saygm-claude
  base_url: https://api.saygm.com
  default: claude-opus-5-5
  2. 2~/.hermes/.env
    bash
    SAYGM_API_KEY=<YOUR_SAYGM_KEY>
  3. 3Run
    bash
    hermes

Switch models with /model, which lists the provider's models, or name one: /model custom:saygm-claude:claude-sonnet-5.

First request

  1. 1Send a first request
    bash
    hermes -z "say gm"

Setup for GPT and open models

Use this setup with the models listed under GPT, Open models and Private (TEE) below.

  1. 1~/.hermes/config.yaml
    yaml
    providers:
  saygm:
    base_url: https://api.saygm.com/v1
    key_env: SAYGM_API_KEY
    api_mode: chat_completions
    models:
      - gpt-6-luna
      - gpt-6-astra
      - kimi-k3-tee

model:
  provider: custom:saygm
  base_url: https://api.saygm.com/v1
  default: gpt-6-luna
  2. 2~/.hermes/.env
    bash
    SAYGM_API_KEY=<YOUR_SAYGM_KEY>
  3. 3Run
    bash
    hermes

Switch models with /model, which lists the provider's models, or name one: /model custom:saygm:gpt-6-astra.

First request

  1. 1Send a first request
    bash
    hermes -z "say gm"

Troubleshooting

Why are there two base URLs?
The path depends on api_mode: anthropic_messages adds /v1/messages itself, so use the API root; chat_completions needs the base URL ending in /v1. Keep each model under a provider matching its api_shapes in saygm models --json.
Setting a GLM-5.3-Flash reasoning effort.
On gateways with reasoning-effort normalization, minimal and medium map to GLM-5.3-Flash's native low level, xhigh maps to max, and low, high, and max pass through unchanged. Hermes sends medium by default, which the gateway maps to low. No manual Hermes override is needed for medium after the gateway update. To choose a native level explicitly, set agent.reasoning_overrides for the exact model ID, for example: agent: { reasoning_overrides: { glm-5.3-flash: high, glm-5.3-flash-tee: high } }. GLM-5.3-Flash always reasons, so none and disable are unsupported.
402 on the first request.
Check the key's spendable amount with saygm balance and add credit at saygm.com/dashboard/credits. A request's required balance depends on the model and token limits.
Adding another model.
Add its id under models in the provider, then switch to it with /model.
Hermes' docs spell the keys api and transport.
Hermes accepts both spellings. base_url and api_mode match its current example config.
Hermes picks the wrong context length.
Add context_length under the provider with the model's real limit.
401 Unauthorized.
Hermes reads SAYGM_API_KEY from ~/.hermes/.env. Check the variable name matches key_env.

Models you can use with Hermes Agent

Hermes runs SayGM's Claude, GPT and open models, through providers matched to each model’s API format.

Claude

11 models
  • Claude Opus 5.5claude-opus-5-5Best qualityThe newest Opus: hard coding and long agentic work at a lower Opus price.In $4.00$3.50Out $20.00$17.51/1M
  • Claude Sonnet 5claude-sonnet-5Best valueEveryday coding and agent work at a fraction of Opus pricing.In $2.00$1.75Out $10.00$8.76/1M
  • Claude Haiku 5.5claude-haiku-5-5Low-cost Claude with adaptive reasoning for quick edits and high-volume tasks.In $0.10$0.09Out $0.50$0.44/1MOffered by miners now
  • Claude Sonnet 5.5claude-sonnet-5-5Anthropic's best combination of speed and intelligence, faster than Sonnet 5.In $2.00$1.75Out $10.00$8.76/1M
  • Claude Fable 5.1claude-fable-5-1Anthropic's most capable model, for the most demanding long-horizon tasks.In $10.00$8.76Out $50.00$43.78/1M
  • Claude Fable 5claude-fable-5Anthropic's top tier before Fable 5.1, for deep reasoning and agentic work.In $10.00$8.76Out $50.00$43.78/1M
  • Claude Opus 5claude-opus-5Opus 5 for complex coding and agents that need careful reasoning.In $5.00$4.38Out $25.00$21.89/1M
  • Claude Opus 4.8claude-opus-4-8Opus 4.8 for teams whose prompts are tuned to the Opus 4 line.In $5.00$4.38Out $25.00$21.89/1M
  • Claude Opus 4.7claude-opus-4-7Opus 4.7 for workflows pinned to that release.In $5.00Out $25.00/1M
  • Claude Sonnet 4.6claude-sonnet-4-6Sonnet 4.6 for workflows pinned to the previous Sonnet.In $3.00$2.63Out $15.00$13.13/1M
  • Claude Haiku 4.5claude-haiku-4-5Fast, low-cost Claude for quick edits, subagents and high-volume calls.In $1.00Out $5.00/1M

GPT

14 models
  • GPT-6 Astragpt-6-astraBest qualityOpenAI's top GPT-6 model, for the hardest coding and reasoning work.In $10.00$5.67Out $50.00$28.33/1M
  • GPT-6 Lunagpt-6-lunaBest valueLow-cost GPT-6 for fast coding loops and high-volume agent turns.In $0.10$0.06Out $0.50$0.28/1M
  • GPT-6.1 Solgpt-6.1-solNear-Astra GPT-6 for complex coding and computer use at a lower price.In $2.00$1.13Out $10.00$5.67/1M
  • GPT-6.1 Terragpt-6.1-terraFast GPT-6.1 for visual reasoning and computer use.In $1.00$0.61Out $5.00$3.04/1M
  • GPT-6 Solgpt-6-solMid-priced GPT-6 for day-to-day coding with strong reasoning.In $2.00$1.13Out $10.00$5.67/1M
  • GPT-5.6 Solgpt-5.6-solThe larger GPT-5.6 model, for demanding coding and analysis.In $4.00$2.27Out $20.00$11.33/1M
  • GPT-5.6 Terragpt-5.6-terraMid-sized GPT-5.6 balancing capability and cost.In $2.00$1.13Out $12.00$6.80/1M
  • GPT-5.6 Lunagpt-5.6-lunaLow-cost GPT-5.6 for quick, high-volume requests.In $0.20$0.11Out $1.20$0.68/1M
  • GPT-5.5gpt-5.5GPT-5.5 reasoning model for workflows tuned to it.In $5.00$4.38Out $30.00$26.27/1M
  • GPT-5.4gpt-5.4GPT-5.4 for general coding and writing at a moderate price.In $2.50$2.19Out $15.00$13.13/1MOffered by miners now
  • GPT-5.4 minigpt-5.4-miniSmall GPT-5.4 for fast, inexpensive everyday tasks.In $0.75$0.66Out $4.50$3.94/1MOffered by miners now
  • GPT-5.4 nanogpt-5.4-nanoThe smallest GPT-5.4, for classification and simple high-volume calls.In $0.20$0.18Out $1.25$1.09/1M
  • o3OpenAI's o3 reasoning model for multi-step math, science and code.In $2.00$1.75Out $8.00$7.00/1MOffered by miners now
  • o4-miniCompact o-series reasoning at a low price.In up to $1.10Out up to $4.40/1M

Open models

12 models
  • gpt-oss-20bOpenAI's small open-weight model for cheap, simple tasks.In $0.07$0.05Out $0.30$0.21/1M
  • Kimi K3kimi-k3Kimi K3, an open-weight reasoning model built for coding agents.In $3.00$1.72Out $15.00$8.61/1M
  • Kimi K3.1 Prokimi-k3.1-proKimi K3.1 Pro for coding, 3D creation and complex financial research.In $2.00$1.22Out $10.00$6.08/1M
  • GLM-5.3glm-5.3Z.ai's GLM-5.3 open-weight model for coding and agent tasks.In $1.40$0.81Out $4.40$2.54/1M
  • GLM-5.3-Flashglm-5.3-flashThe fast, low-cost GLM-5.3 for quick coding help.In $0.15$0.12Out $0.50$0.40/1M
  • GLM-5.2glm-5.2GLM-5.2 open weights for coding and reasoning.In $1.40$0.58Out $4.40$1.83/1M
  • DeepSeek-V4.1-Flashdeepseek-v4.1-flashDeepSeek's fast V4.1 model for low-cost coding and chat.In $0.30$0.04Out $1.20$0.16/1M
  • DeepSeek-V4.1-Flash-1017deepseek-v4.1-flash-1017DeepSeek's fast V4.1 model for low-cost coding and chat.In $0.30$0.06Out $1.20$0.22/1M
  • DeepSeek-V4-Flash-0731deepseek-v4-flash-0731DeepSeek V4 Flash for inexpensive, quick coding help.In $0.44$0.07Out $1.32$0.20/1M
  • Qwen3.8-27Bqwen3.8-27bQwen3.8 27B, a compact open-weight model for everyday coding.In $0.50$0.05Out $3.00$0.29/1M
  • Qwen3.6-35B-A3Bqwen3.6-35b-a3bQwen3.6 35B mixture-of-experts, fast and cheap for light tasks.In $0.248$0.074Out $1.485$0.444/1MOffered by miners now
  • MiMo-V2.6-Pro-UltraSpeedmimo-v2.6-pro-ultraspeedXiaomi MiMo V2.6 Pro tuned for speed, with a 1M-token context.In $4.35$3.81Out $8.70$7.62/1M

Private (TEE)

16 models
  • Kimi-K3 (TEE)kimi-k3-teePrivate (TEE)Kimi K3, an open-weight coding and agent model, run inside a TEE.In $3.00$2.51Out $15.00$12.56/1MOffered by miners now
  • Kimi K3.1 Pro (TEE)kimi-k3.1-pro-teeKimi K3.1 Pro run inside a TEE, for coding and research on private data.In $2.00$1.75Out $10.00$8.76/1M
  • Kimi-K2.6 (TEE)kimi-k2.6-teeKimi K2.6 open weights inside a TEE, for private agent work.In $0.50$0.44Out $2.85$2.48/1MOffered by miners now
  • GLM-5.3 (TEE)glm-5.3-teeGLM-5.3 run inside a TEE, for coding and agent tasks on private code.In $1.40$1.23Out $4.40$3.85/1M
  • GLM-5.3-Flash (TEE)glm-5.3-flash-teeThe fast, low-cost GLM-5.3 inside a TEE, for quick private coding help.In $0.15Out $0.50$0.49/1M
  • GLM-5.2 (TEE)glm-5.2-teeGLM-5.2 run inside a TEE, for coding on private code.In $1.40$1.23Out $4.40$3.85/1M
  • GLM-5.1 (TEE)glm-5.1-teeGLM-5.1 run inside a TEE, for private reasoning tasks.In $1.40$1.19Out $4.40$3.74/1MOffered by miners now
  • DeepSeek-V4.1-Flash (TEE)deepseek-v4.1-flash-teeDeepSeek V4.1 Flash inside a TEE, for low-cost private work.In $0.30$0.26Out $1.20$1.05/1MOffered by miners now
  • DeepSeek-V4.1-Flash-1017 (TEE)deepseek-v4.1-flash-1017-teeDeepSeek V4.1 Flash inside a TEE, for low-cost private work.In $0.30$0.28Out $1.20$1.11/1M
  • DeepSeek-V4-Flash-0731 (TEE)deepseek-v4-flash-0731-teeDeepSeek V4 Flash inside a TEE, for low-cost private work.In $0.44$0.43Out $1.32$1.29/1MOffered by miners now
  • DeepSeek-V3.2 (TEE)deepseek-v3.2-teeDeepSeek V3.2 inside a TEE, for private general-purpose tasks.In $1.00$0.98Out $1.00$0.98/1M
  • Qwen3.8-27B (TEE)qwen3.8-27b-teeQwen3.8 27B with private prompts inside a NEAR-attested TEE.In $0.44$0.19Out $3.30$1.46/1MOffered by miners now
  • Qwen3.6-27B (TEE)qwen3.6-27b-teeQwen3.6 27B inside a TEE, for private everyday tasks.In $0.30$0.29Out $2.00$1.96/1MOffered by miners now
  • Qwen3.5-397B-A17B (TEE)qwen3.5-397b-a17b-teeQwen3.5 397B, a large open-weight model run inside a TEE.In $0.45$0.44Out $3.00$2.94/1MOffered by miners now
  • MiMo-V2.6-Pro-UltraSpeed (TEE)mimo-v2.6-pro-ultraspeed-teeXiaomi MiMo V2.6 Pro tuned for speed, run inside a TEE.In $4.35$3.81Out $8.70$7.62/1MOffered by miners now
  • Gemma 4 31B turbo (TEE)gemma-4-31b-turbo-teeGoogle's Gemma 4 31B open weights inside a TEE, low cost and private.In $0.12Out $0.37$0.36/1M

About Hermes Agent

Hermes Agent is Nous Research's open-source, self-improving agent. It runs in your terminal or on a server you reach from Telegram, Discord, Slack and other chat apps, turns what it learns into reusable skills, and keeps memory across sessions. People use it as a long-running personal agent for scheduled jobs and everyday tasks, on any model provider.