All models in one place.
Claude, GPT, Gemini and open models, all through one router, in the desktop apps your teams already use.
| Model | Cost relative |
|---|---|
| Claude Fable 5 | |
| Claude Opus 5 | |
| Claude Opus 4.8 | |
| Claude Opus 4.7 | |
| Claude Sonnet 5 | |
| Claude Sonnet 4.6 | |
| Claude Haiku 4.5 | |
| GPT 5.6 Sol | |
| GPT 5.6 Terra | |
| GPT 5.6 Luna | |
| GPT 5.5 | |
| GPT 5.4 | |
| GPT 5.4 Mini | |
| GPT 5.4 Nano | |
| Gemini 3.1 Pro (preview) | |
| Gemini 3.7 Flash | |
| Gemini 3.6 Flash | |
| Gemini 3.5 Flash | |
| Gemini 3 Flash (preview) | |
| Gemini 3.1 Flash Lite | |
| Kimi K3open | |
| Kimi K2.7 Codeopen | |
| Kimi K2.6open | |
| DeepSeek V4 Proopen | |
| DeepSeek V4 Flashopen | |
| GLM 5.2open |
As of 11 September 2026. Marks show relative list price. Fewer marks means more volume for the same spend. Not a quote.
Your keys or ours. Same models either way.
Bring your own keys
Keep your Anthropic, OpenAI and Google relationships, your rates and any committed-spend discounts.
Harriet managed credits
A single invoice with access to all the frontier labs as well as high-quality open-weight models.
What teams ask about models.
Can we keep the Claude or ChatGPT models we already use?
Yes. Bring your own Anthropic, OpenAI or Google keys and keep your existing provider relationship, rates and committed-spend discounts, or use Harriet managed credits for a single invoice with access to all the frontier labs as well as high-quality open-weight models. Either way the models are the same.
Where do the models run?
Harriet's gateway forwards prompts to EU-region endpoints, so Claude, OpenAI and Gemini are processed in the EU and covered by your providers' EU data processing terms. Open models are hosted in the EU or US, your choice.
How do we switch a team to a different model?
It is a setting on the team profile. Change the default model and it applies to everyone in that team on their next request. No reinstall, no retraining, no migration.
A model we need is not listed. What then?
Tell us which one. The router is built to add providers and models as they ship, and this roster moves often; the date under the catalogue says how current it is.
Pick the model per team, not per company.
Book 20 minutes and we will map your teams to the models that fit, on your keys or ours.