en-models.md
OmniaKey routes to a curated catalog from four vendors as peers — use the model id below in any compatible SDK or coding agent. These models support tool use; those marked fast are tuned for low latency. See the API Reference for how to call these ids over each protocol.
| Model ID | Context | Notes |
|---|---|---|
claude-opus-4-8 | 1M | Flagship. Vision, tool use |
claude-opus-4-7 | 1M | Vision, tool use |
claude-opus-4-6 | 1M | Vision, tool use |
claude-sonnet-5 | 1M | Balanced. Vision, tool use |
claude-sonnet-4-6 | 1M | Balanced. Vision, tool use |
claude-haiku-4-5 | 200K | Fast. Vision, tool use |
| Model ID | Context | Notes |
|---|---|---|
gpt-5.5 | 1M | Flagship. Vision, tool use |
gpt-5.4 | 1M | Vision, tool use |
gpt-5.4-mini | 1M | Fast. Vision, tool use |
| Model ID | Context | Notes |
|---|---|---|
gemini-3.1-pro-preview | 1M | Flagship. Vision, tool use |
gemini-3-flash-preview | 1M | Fast. Vision, tool use |
gemini-2.5-pro | 1M | Vision, tool use |
gemini-2.5-flash | 1M | Fast. Vision, tool use |
| Model ID | Context | Notes |
|---|---|---|
grok-4.5 | 500K | Flagship coding model. Vision, tool use |
You pay per token against a prepaid balance — no monthly fee, no tiers. Launch pricing differs by provider and is shown line by line on the pricing page; live rates are also shown in the Dashboard.
The lineup grows over time. The pricing page and Dashboard always reflect what’s currently available.
Gemini-native APIErrors and Rate Limits
⌘I