Models and prices
This page has the model ids for the model field and explains the columns of the price list. The prices themselves are on the pricing page, kept in one place so they cannot drift.
Model list#
The list as of . In a request use the value of the Id column letter for letter.
| Model | Id | Alias | Protocols | Context | Max output |
|---|---|---|---|---|---|
| GPT-6 Astra | gpt-6-astra | — | Chat, Codex, Claude Code | 1.05M | 128K |
| GPT-6.1 Sol | gpt-6.1-sol | — | Chat, Codex, Claude Code | 1.05M | 128K |
| GPT-6.1 Terra | gpt-6.1-terra | — | Chat, Codex, Claude Code | 1.05M | 128K |
| GPT-6 Sol | gpt-6-sol | — | Chat, Codex, Claude Code | 1.05M | 128K |
| GPT-6 Luna | gpt-6-luna | — | Chat, Codex, Claude Code | 1.05M | 128K |
| GPT-5.6 Sol | gpt-5.6-sol | — | Chat, Codex, Claude Code | 1.05M | 128K |
| GPT-5.6 Terra | gpt-5.6-terra | — | Chat, Codex, Claude Code | 1.05M | 128K |
| GPT-5.6 Luna | gpt-5.6-luna | — | Chat, Codex, Claude Code | 1.05M | 128K |
| GPT-5.5 | gpt-5.5 | — | Chat, Codex, Claude Code | 272K | 128K |
| GPT-5.4 | gpt-5.4 | — | Chat | — | 128K |
| Claude Opus 5.5 | claude-opus-5-5 | opus | Chat, Codex, Claude Code | 1M | 64K |
| Claude Opus 5 | claude-opus-5 | — | Chat, Codex, Claude Code | 1M | 64K |
| Claude Opus 4.8 | claude-opus-4-8 | — | Chat, Codex, Claude Code | 1M | 64K |
| Claude Opus 4.7 | claude-opus-4-7 | — | Chat, Codex, Claude Code | 1M | 64K |
| Claude Opus 4.6 | claude-opus-4-6 | — | Chat, Codex, Claude Code | 1M | 64K |
| Claude Sonnet 5.5 | claude-sonnet-5-5 | sonnet | Chat, Codex, Claude Code | 1M | 64K |
| Claude Sonnet 5 | claude-sonnet-5 | — | Chat, Codex, Claude Code | 1M | 64K |
| Claude Sonnet 4.6 | claude-sonnet-4-6 | — | Chat, Codex, Claude Code | 1M | 64K |
| Claude Haiku 5.5 | claude-haiku-5-5 | haiku | Chat, Codex, Claude Code | — | 32K |
| Claude Haiku 4.5 | claude-haiku-4-5 | — | Chat, Codex, Claude Code | — | 32K |
| Claude Fable 5.1 | claude-fable-5-1 | fable | Chat, Codex, Claude Code | 1M | 64K |
| Claude Fable 5 | claude-fable-5 | — | Chat, Codex, Claude Code | — | 64K |
| Gemini 4 Argon | gemini-4-argon | — | Chat, Codex, Claude Code | 1.05M | 65.5K |
| Gemini 3.8 Flash | gemini-3.8-flash | — | Chat, Codex, Claude Code | 1.05M | 65.5K |
| Gemini 3.1 Pro | gemini-3.1-pro | — | Chat | 1M | 65.5K |
| Kimi K3.1 Pro | kimi-k3.1-pro | — | Chat, Codex, Claude Code | 1.05M | 32.8K |
| Kimi K3 | kimi-k3 | — | Chat, Codex, Claude Code | 1.05M | 32.8K |
| DeepSeek V4.1 Flash | deepseek-v4.1-flash | — | Chat, Codex, Claude Code | 1.05M | 32.8K |
| DeepSeek V4 Pro | deepseek-v4-pro | — | Chat, Codex, Claude Code | 1M | 32.8K |
| GLM 5.3 | glm-5.3 | — | Chat, Codex, Claude Code | 1M | 32.8K |
| Grok 4.7 | grok-4.7 | — | Chat, Codex, Claude Code | 500K | 32.8K |
| Grok 4.6 | grok-4.6 | — | Chat, Codex, Claude Code | 500K | 32.8K |
- Context and maximum output are in tokens. A dash under Context means we have no window size on record, not that there is none.
- The maximum output is also how much output we reserve on the balance when a request has no
max_tokens. - An alias is a short name Claude Code sends instead of the full id; we resolve it on our side.
The API itself gives the list for your key and protocol; a key restricted to some models sees only those:
curl "https://api.aicreator.app/v1/models?surface=chat" \
-H "Authorization: Bearer $AIC_API_KEY"The Protocols column#
A model is available only on the protocols it lists. On the others a request for it returns 400 model_not_available.
| Tag | Protocol | Used by | ?surface= |
|---|---|---|---|
| Claude Code | Anthropic Messages | Claude Code, Anthropic SDK | anthropic |
| Codex | OpenAI Responses | Codex CLI, OpenAI SDK | responses |
| Chat | OpenAI Chat Completions | Cline, OpenCode, Aider, OpenAI SDK | chat |
The price table’s columns#
Every price is in dollars per million tokens. A request is charged on four counters, each at its own price:
- Input — the tokens of the request that were neither read from the cache nor written to it.
- Cache read — the tokens of the request the provider took from the prompt cache. The price is set per model; there is no single share of the input price.
- Cache write 5 min / 1 h — the tokens the request put into the cache. There are two prices: for a cache that lives 5 minutes and for one that lives an hour.
- Output — the tokens of the model’s answer.
A dash in a cache column means the model has no such price. Cache reads and writes are counted apart from input: a token lands in one counter only.
Claude Code and Codex cache the prompt themselves, so in a long session most input goes at the cache read price.
Official and Saving#
Official is the model vendor’s own input and output price; the sources are listed under the price table. Saving is computed from our price and the official one when the site is built; nobody types it. It is the smallest of the differences on input, output and cache reads, rounded down. A dash means the model has no sourced official price, or at least one of those prices of ours is not lower.