# Models and pricing

> Discover exact model IDs, current availability, and prices per one million tokens.

The runtime source of truth is the authenticated model catalog. For an OpenAI-compatible client, call `GET /v1/models` with an API-credit key or `GET /token/v1/models` with an API-token key. It lists only models with active pricing and a configured route.

This page is the API reference for model ids. To browse every model with its rate in your currency, see the [model catalogue on scalemax.pro](https://scalemax.pro/models).

## Model ids and versions

Claude ids ending in \[1m\] are the same model with the full 1M-token context — use either id. Ids ending in `-v2`, `-v3`, `-new` or `-preview` are other versions of the same model, each with its own price; the plain id is the one to start with.

## Provider families

Public families: OpenAI, Claude, Gemini, Grok, Qwen, DeepSeek, MiMo, and Local Models. Token checkout sells provider-scoped packs for Claude, OpenAI, combined Claude + OpenAI, Gemini, Grok, Qwen, and Local Models, plus exact-model packs such as DeepSeek V4 Flash and GLM 5.2. MiMo ids are sold inside the Local Models pack.

## List live models

GET `$SCALEMAX_BASE_URL/models`

Set `SCALEMAX_BASE_URL` to `https://api.scalemax.pro/v1` for API credit or `https://api.scalemax.pro/token/v1` for API tokens. Native Anthropic clients discover models through their origin plus `/v1/models`.

**macOS / Linux**

```bash
curl "$SCALEMAX_BASE_URL/models" \
  -H "Authorization: Bearer YOUR_SCALEMAX_API_KEY"
```

**Windows PowerShell**

```powershell
curl.exe "$env:SCALEMAX_BASE_URL/models" `
  -H "Authorization: Bearer YOUR_SCALEMAX_API_KEY"
```

## Local Models

The Local Models token pack covers 41 public ids, grouped by maker. `glm-5.3` is priced but not yet callable. Withdrawal from this list is permanent for ids the upstream no longer serves; do not reuse a missing id.

### DeepSeek

- `deepseek-v4-flash`
- `deepseek-v4-flash-new`
- `dsv4-flash-0731-unlimited`
- `claude-sonnet-4-6`
- `deepseek-v4-pro`
- `deepseek-v3.2`

### Gemini

- `gemini-3-flash-preview`
- `gemini-3.1-flash-lite-preview`
- `gemini-3.5-flash-lite`
- `gemini-3.8-flash`

### Grok

- `grok-4.5`
- `grok-4.6`
- `grok-build`

### GLM

- `glm-5`
- `glm-5.1`
- `glm-5.2`
- `glm-5.3` — coming soon (priced, not yet routed)
- `glm-5.3-flash`
- `glm-5.2-v2`
- `glm-5.3-v2`
- `glm-5.3-flash-v2`
- `glm-5.3-flash-v3`

### MiniMax

- `minimax-m2.5`
- `minimax-m2.7`
- `minimax-m3`

### Kimi

- `kimi-k2.6`
- `kimi-k2.7-code`
- `kimi-k3`

### Hunyuan

- `hunyuan-3`
- `hy4`
- `hy4-preview`

### MiMo

- `mimo-v2.5-pro`

### Mistral

- `mistral-large-3`

### Nemotron

- `nemotron-lightning-3.5`

### Qwen

- `qwen-3.7-flash`
- `qwen-3.7-max`
- `qwen-3.8-27b`
- `qwen3.8-flash`
- `qwen3.8-max`

### Muse

- `muse-spark-1.2`

### OpenAI

- `gpt-oss-120b`

## Current pricing

Prices are customer-facing USD rates per 1M tokens for representative catalog ids. GPT-5.6 Sol/Terra, GPT-6 Sol/Luna/Astra, GPT-6.1 Sol, and Fable rows use the current ScaleMax standard rate; GPT-5.6 Luna uses the Luna card. Other rows show the current ScaleMax configured rate. Uncached input and cache writes bill at the published input rate unless a separate public rate is configured. Provider-reported cached-read input bills at the model’s published cached-input rate where one is configured; otherwise it bills at that model’s ordinary input rate. The ledger split behind that rule is described in [Pricing and accounting architecture](https://docs.scalemax.pro/docs/pricing-architecture), and pack prices are listed in [Billing, API credits & token packs](https://docs.scalemax.pro/docs/billing).

Swipe horizontally to compare every price column.

Current ScaleMax customer model prices

| Model | Maker | Input | Output | Source |
| --- | --- | --- | --- | --- |
| `claude-haiku-4-5-20251001` | Anthropic | $1.00 | $5.00 | ScaleMax configured rate |
| `claude-haiku-5-5` | Anthropic | $0.20 | $1.00 | ScaleMax configured rate |
| `claude-opus-4-6` | Anthropic | $5.00 | $25.00 | ScaleMax configured rate |
| `claude-opus-4-7` | Anthropic | $5.00 | $25.00 | ScaleMax configured rate |
| `claude-opus-4-8` | Anthropic | $5.00 | $25.00 | ScaleMax configured rate |
| `claude-opus-5` | Anthropic | $5.00 | $25.00 | Anthropic official rate |
| `claude-opus-5-5` | Anthropic | $5.00 | $25.00 | ScaleMax configured rate |
| `claude-sonnet-5` | Anthropic | $2.00 | $10.00 | ScaleMax configured rate |
| `claude-sonnet-5-5` | Anthropic | $2.00 | $10.00 | ScaleMax configured rate |
| `fable-5` | Anthropic | $10.00 | $50.00 | Anthropic-family Fable pricing |
| `gpt-5.6-luna` | OpenAI | $0.522 | $3.132 | ScaleMax configured rate |
| `gpt-5.6-sol` | OpenAI | $5.00 | $30.00 | ScaleMax standard |
| `gpt-5.6-terra` | OpenAI | $2.00 | $12.00 | ScaleMax standard |
| `gpt-6-astra` | OpenAI | $15.00 | $60.00 | ScaleMax standard |
| `gpt-6-luna` | OpenAI | $0.10 | $0.50 | ScaleMax standard |
| `gpt-6-sol` | OpenAI | $2.00 | $10.00 | ScaleMax standard |
| `gpt-6.1-sol` | OpenAI | $2.00 | $10.00 | ScaleMax standard |
| `gemini-3-flash-preview` | Gemini | $0.50 | $3.00 | ScaleMax configured rate |
| `gemini-3.1-flash-lite-preview` | Gemini | $0.25 | $1.50 | ScaleMax configured rate |
| `grok-4.5` | Grok | $2.00 | $6.00 | ScaleMax configured rate |
| `qwen3.6-plus` | Qwen | $2.00 | $12.00 | ScaleMax configured rate |
| `qwen3.7-plus` | Qwen | $2.00 | $8.00 | ScaleMax configured rate |
| `gemini-3.5-flash-lite` | Gemini | $0.25 | $1.50 | ScaleMax configured rate |
| `gemini-3.8-flash` | Gemini | $0.75 | $3.75 | ScaleMax configured rate |
| `deepseek-v4-flash` | DeepSeek | $0.19 | $0.51 | ScaleMax configured rate |
| `glm-5.2` | GLM | $1.925 | $3.828 | ScaleMax configured rate |

## Cached input rates

These models publish their own cached-read rate. The rate covers provider-reported cached-read input only: cache writes and uncached input bill at the model’s ordinary input rate unless a separate public write rate is configured, and the cached rate does not change with context length.

Swipe horizontally to see the cached rate column.

Customer cached-read input rates per one million tokens

| Model | Cached input |
| --- | --- |
| `gpt-5.6-luna` | $0.0522 |
| `gpt-6-sol` | $0.20 |
| `gpt-6-luna` | $0.01 |
| `gpt-6.1-sol` | $0.10 |
| `gemini-3-flash-preview` | $0.050 |
| `gemini-3.1-flash-lite-preview` | $0.025 |
| `gemini-3.5-flash-lite` | $0.025 |
| `grok-4.5` | $0.500 |
| `qwen3.6-plus` | $0.200 |
| `qwen3.7-plus` | $0.400 |

## GPT long-context pricing

When a request contains more than 272K input tokens, long-context rates apply to the entire request—not only the tokens above the threshold. GPT-5.6 Luna uses the Luna card’s 2× / 1.5× tier. Sol and Terra keep their published long-context rows.

Swipe horizontally to compare long-context rates.

GPT long-context customer rates

| Model | Input | Output |
| --- | --- | --- |
| `gpt-5.6-sol` | $10.00 | $45.00 |
| `gpt-5.6-terra` | $4.00 | $18.00 |
| `gpt-5.6-luna` | $1.044 | $4.698 |

Above the threshold, uncached input and output move to the long-context rates; a published cached-input rate is flat and does not change with context length. Internal provider cache prices remain separate from customer billing. Fable standard rows, GPT-6 Sol/Luna and GPT-6.1 Sol above have no long-context tier.

## Embeddings

POST `/v1/embeddings` is OpenAI-compatible. Send `input` as a string, an array of up to 2,048 strings, or token-id arrays; `dimensions` (text-embedding-3 models only), `encoding_format` (`float` or `base64`) and `user` are passed through. Embeddings bill on input tokens only (the `usage.prompt_tokens` in the response); there is no output charge. API-credit keys pay the rate below. OpenAI and combined token packs are debited in proportion to that rate, so a pack unit buys the same dollar value of embedding as of any other model. The ids are listed by `GET /v1/models` with `"family": "embedding"` and are not accepted by the chat endpoints.

Swipe horizontally to compare every column.

Embeddings models, dimensions and prices per one million input tokens

| Model | Dimensions | Max input | Input / 1M |
| --- | --- | --- | --- |
| `text-embedding-3-small` | 1,536 (shortenable) | 8,191 tokens | $0.02 |
| `text-embedding-3-large` | 3,072 (shortenable) | 8,191 tokens | $0.13 |
| `text-embedding-ada-002` | 1,536 | 8,191 tokens | $0.10 |

**macOS / Linux**

```bash
curl "https://api.scalemax.pro/v1/embeddings" \
  -H "Authorization: Bearer $SCALEMAX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"text-embedding-3-small","input":"hello"}'
```

**Windows PowerShell**

```powershell
@'
{"model":"text-embedding-3-small","input":"hello"}
'@ | Set-Content -Path body.json

curl.exe "https://api.scalemax.pro/v1/embeddings" `
  -H "Authorization: Bearer $env:SCALEMAX_API_KEY" `
  -H "Content-Type: application/json" `
  --data-binary "@body.json"
```

PowerShell changes quotes inside a JSON argument, so the request body goes into body.json first and curl.exe sends that file. Type curl.exe, not curl: in Windows PowerShell 5.1, curl is a different command.

## Choose a model

Use the exact IDs returned by the API. Capability details can vary by provider route, so test the model in your workload before a large rollout.