Models and pricing.
Choose a model by its capabilities and compare prices for the same number of tokens.
Found: 32 of 32
Reset filtersApproximate tokens per $1 of balance based on observed usage. Results depend on the mix of input, output and cache.
Model names are pinned. Scroll the price columns to the right →
Context and cache prices: expand the model with +.
| Model / Model ID | Context | InputInput tokens | OutputOutput tokens | Cached inputReading cache | Cache writeWriting cache |
|---|---|---|---|---|---|
GPT-6.2 AstraOpenAI* ≈ 41 million tokens per $1gpt-6.2-astra | 1,050,000tokens | $0.17$10 | $0.85$50 | $0.017$1 | $0$12.5 |
GPT-6.2 Astra · OpenAI A model for complex agentic and coding tasks with up to 1.05M-token context. Model context: 1,050,000 tokens
Official prices only: the row shows the base rate up to 272,000 tokens of context; above that — ×2, according to the source. This multiplier does not apply to our prices. Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
GPT-6.2 SolOpenAI* ≈ 73 million tokens per $1gpt-6.2-sol | 1,050,000tokens | $0.1$2 | $0.5$10 | $0.01$0.2 | $0$2.5 |
GPT-6.2 Sol · OpenAI A model for coding, analysis, and agentic tasks. Model context: 1,050,000 tokens
Official prices only: the row shows the base rate up to 272,000 tokens of context; above that — ×2, according to the source. This multiplier does not apply to our prices. Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
GPT-6.1 TerraOpenAI ≈ 114 million tokens per $1gpt-6.1-terra | 1,050,000tokens | $0.06$1 | $0.3$5 | $0.006$0.05 | $0$1.25 |
GPT-6.1 Terra · OpenAI A fast model for visual reasoning, computer use, and writing. The row shows official Standard prices up to 272000 input tokens. Above 272000: input and cache rates ×2, output ×1.5 for the full request. Official cache writes per 1M tokens: $1.25 for 30 minutes of retention, $2.50 for 12 hours. These conditions do not change Etage prices. Model context: 1,050,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
GPT-6 LunaOpenAI ≈ 200 million tokens per $1gpt-6-luna | 1,050,000tokens | $0.036$0.2 | $0.18$0.75 | $0.0036$0.02 | $0$0.25 |
GPT-6 Luna · OpenAI A model with up to 1.05 million tokens of context. Model capabilities have not been specified yet. Model context: 1,050,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
GPT Image 2.5 FlareOpenAI 50 images per $1gpt-image-2.5-flare | Image — $0.02 / per generation | ||||
GPT Image 2.5 Flare · OpenAI Image generation with a fixed price per image. $0.02 / per generation | |||||
GPT Image 2.5 SunburstOpenAI 40 images per $1gpt-image-2.5-sunburst | Image — $0.025 / per generation | ||||
GPT Image 2.5 Sunburst · OpenAI Image generation with a fixed price per image. $0.025 / per generation | |||||
Kimi K3.1 ProMoonshot AI ≈ 48 million tokens per $1kimi-k3.1-pro | 1,048,576tokens | $0.14$2 | $0.7$10 | $0.014$0.2 | $0$2 |
Kimi K3.1 Pro · Moonshot AI A model for coding, analysis, and agentic tasks. The official Cache write price in the table covers 5 minutes of storage; 1 hour costs $4 per 1M tokens. Model context: 1,048,576 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Kimi K3Moonshot AIDiscount ≈ 68 million tokens per $1kimi-k3 | 1,048,576tokens | $0.1$3 | $0.5$15 | $0.01$0.3 | $0$3 |
Kimi K3 · Moonshot AI A model for coding, analysis, and agentic tasks. The official Cache write price in the table covers 5 minutes of storage; 1 hour costs $6 per 1M tokens. Model context: 1,048,576 tokens
Regular Etage rates · USD per 1 M tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Claude Fable 5.1Anthropic ≈ 9 million tokens per $1claude-fable-5-1 | 1,000,000tokens | $0.7$10 | $3.5$50 | $0.07$0.25 | $0$12.5 |
Claude Fable 5.1 · Anthropic A Claude model for coding, analysis, and agentic tasks. The official Cache write price in the table covers 5 minutes of storage; 1 hour costs $20 per 1M tokens. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Claude Fable 5.2Anthropic ≈ 9 million tokens per $1claude-fable-5-2 | 1,000,000tokens | $0.7$8 | $3.5$40 | $0.07$0.2 | $0$10 |
Claude Fable 5.2 · Anthropic A Claude model for coding, analysis, and agentic tasks. The official Cache write price in the table covers 5 minutes of storage. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Claude Opus 5.5Anthropic ≈ 19 million tokens per $1claude-opus-5-5 | 1,000,000tokens | $0.35$4 | $1.75$20 | $0.035$0.2 | $0$5 |
Claude Opus 5.5 · Anthropic A Claude model for coding, analysis, and agentic tasks. The official Cache write price in the table covers 5 minutes of storage; 1 hour costs $8 per 1M tokens. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Claude Sonnet 5.5Anthropic ≈ 32 million tokens per $1claude-sonnet-5-5 | 1,000,000tokens | $0.21$2 | $1.05$10 | $0.021$0.1 | $0$2.5 |
Claude Sonnet 5.5 · Anthropic A Claude model for coding, analysis, and agentic tasks. The official Cache write price in the table covers 5 minutes of storage; 1 hour costs $4 per 1M tokens. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Claude Haiku 5.5Anthropic* ≈ 105 million tokens per $1claude-haiku-5-5 | 1,000,000tokens | $0.065$0.1 | $0.32$0.5 | $0.0065$0.01 | $0$0.125 |
Claude Haiku 5.5 · Anthropic A Claude model for coding, analysis, and agentic tasks. The official Cache write price in the table covers 5 minutes of storage; 1 hour costs $0.2 per 1M tokens. Official base rates cover prompts up to 100000 input tokens, including cache; over 100000, all official rates for the entire request increase fivefold. Model context: 1,000,000 tokens
Official prices only: the row shows the base rate up to 100,000 tokens of context; above that — ×5, according to the source. This multiplier does not apply to our prices. Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Grok 4.7X AITemporary sale −50%* ≈ 146 million tokens per $1grok-4.7 | 500,000tokens | $0.05$2 | $0.25$6 | $0.005$0.5 | $0No data |
Grok 4.7 · X AI xAI flagship model for code and agentic tasks with a 500K-token context. Model context: 500,000 tokens
Official prices only: the row shows the base rate up to 272,000 tokens of context; above that — ×2, according to the source. This multiplier does not apply to our prices. Regular Etage rates · USD per 1 M tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
GLM 5.3Z.AI ≈ 80 million tokens per $1glm-5.3 | 1,000,000tokens | $0.096$1.4 | $0.288$4.4 | $0.0096$0.26 | $0No data |
GLM 5.3 · Z.AI A model for coding, analysis, and agentic tasks. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
GLM-5.3 FlashZ.AI ≈ 115 million tokens per $1glm-5.3-flash | 1,000,000tokens | $0.049$0.15 | $0.18$0.5 | $0.007$0.03 | $0No data |
GLM-5.3 Flash · Z.AI Fast version of GLM 5.3 with forced deep thinking. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
GLM-5.2Z.AI ≈ 350 million tokens per $1glm-5.2 | 1,000,000tokens | $0.02$1.1 | $0.0656$3.86 | $0.002$0.28 | $0$0.28 |
GLM-5.2 · Z.AI A model for coding, analysis, and agentic tasks. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
DeepSeek V4.1 Flash (latest)DeepSeek ≈ 320 million tokens per $1deepseek-v4.1-flash | 1,000,000tokens | $0.045$0.3 | $0.18$1.2 | $0.0018$0.006 | $0$0.000001 |
DeepSeek V4.1 Flash (latest) · DeepSeek New V4.1 Flash — faster and cheaper than V4 Pro, with vision and 1M context. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
DeepSeek V4 Pro 0813DeepSeek deepseek-v4-pro-0813 | 1,000,000tokens | $0.16No data | $0.64No data | $0.016No data | $0No data |
DeepSeek V4 Pro 0813 · DeepSeek A model for coding, analysis, and agentic tasks. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Minimax M3MiniMax minimax-m3 | 512,000tokens | $0.12No data | $0.48No data | $0.012No data | $0No data |
Minimax M3 · MiniMax A model for coding, analysis, and agentic tasks. Model context: 512,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Qwen4 Max PreviewAlibaba qwen4-max-preview | No data | $0.23No data | $0.68No data | $0.03No data | $0No data |
Qwen4 Max Preview · Alibaba Qwen4 Max Preview for text requests and reasoning. Model context: No data
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Qwen3.8 Max 0902AlibabaDiscount ≈ 73 million tokens per $1qwen3.8-max-0902 | 1,000,000tokens | $0.1$2 | $0.5$6 | $0.01$0.25 | $0No data |
Qwen3.8 Max 0902 · Alibaba Latest Qwen3.8 Max (0902) — for coding, analysis, and agentic tasks. Model context: 1,000,000 tokens
Regular Etage rates · USD per 1 M tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Jev 1.13TypeSafe typesafe/jev-1.13 | 262,000tokens | $0.049$0.042 | —— | —— | —— |
Jev 1.13 · TypeSafe A model for typed and structured decisions. Use POST /jev/alpha/decisions, not Chat Completions. Model context: 262,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Jev 1.14TypeSafe typesafe/jev-1.14 | 262,000tokens | $0.049$0.042 | —— | —— | —— |
Jev 1.14 · TypeSafe A model for typed and structured decisions. Use POST /jev/alpha/decisions, not Chat Completions. Model context: 262,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
Jev LatestAlways the latest version typesafe/jev-latest | 262,000tokens | $0.049$0.042 | —— | —— | —— |
Jev Latest · TypeSafe A model for typed and structured decisions. Use POST /jev/alpha/decisions, not Chat Completions. Always the latest version of Jev. Model context: 262,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
GLM-5.3 AbliteratedZ.AI glm-5.3-abliterated | 1,000,000tokens | $0.78No data | $2.6No data | $0.31No data | No dataNo data |
GLM-5.3 Abliterated · Z.AI Abliterated GLM 5.3 for coding, analysis, and agentic tasks. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
GLM-5.3 Flash AbliteratedZ.AI glm-5.3-flash-abliterated | 1,000,000tokens | $0.24No data | $0.98No data | $0.056No data | No dataNo data |
GLM-5.3 Flash Abliterated · Z.AI Fast abliterated version of GLM 5.3. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
JB Kimi K3.1 ProMoonshot AI ≈ 35 million tokens per $1jb-kimi-k3.1-pro | 1,048,576tokens | $0.2No data | $0.8No data | $0.02No data | $0No data |
JB Kimi K3.1 Pro · Moonshot AI A model for coding, analysis, and agentic tasks. Model context: 1,048,576 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
JB Kimi K3Moonshot AI ≈ 35 million tokens per $1jb-kimi-k3 | 1,048,576tokens | $0.2No data | $0.8No data | $0.02No data | $0No data |
JB Kimi K3 · Moonshot AI A model for coding, analysis, and agentic tasks. Model context: 1,048,576 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
JB GLM-5.3 FlashZ.AI ≈ 35 million tokens per $1jb-glm-5.3-flash | 1,000,000tokens | $0.2No data | $0.8No data | $0.02No data | $0No data |
JB GLM-5.3 Flash · Z.AI Fast version of GLM 5.3 with forced deep thinking. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
JB GLM-5.2Z.AI ≈ 35 million tokens per $1jb-glm-5.2 | 1,000,000tokens | $0.2No data | $0.8No data | $0.02No data | $0No data |
JB GLM-5.2 · Z.AI A model for coding, analysis, and agentic tasks. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
JB DeepSeek V4.1 FlashDeepSeek ≈ 35 million tokens per $1jb-deepseek-v4.1-flash | 1,000,000tokens | $0.2No data | $0.8No data | $0.02No data | $0No data |
JB DeepSeek V4.1 Flash · DeepSeek The same V4.1 Flash without censorship. Vision and 1M context. Model context: 1,000,000 tokens
Official* — comparison from the original catalog entries, without independent verification. “No data” does not mean free usage or support for an operation. | |||||
How request costs are calculated
How request costs are calculated
You pay for the amount of data the model receives and generates. This is measured in tokens — small pieces of text. A token can be part of a word, a whole word or punctuation.
The table lists prices in US dollars per 1 million tokens. You do not need to use a million at once: costs are calculated from the actual size of each request.
- Input — what you send to the model
Your message, instructions, conversation history included in the request and document contents. If you send a document and ask for a summary, both the document and your request count as Input. The more data you send, the more input tokens you use.
- Output — what the model generates
The model’s response: text, code, a translation or an analysis. Short answers usually use fewer tokens than detailed ones. Some models also bill reasoning tokens, depending on the model and its pricing.
- Cached input — repeated input costs less
When you keep working in the same chat or session in Claude Code, OpenCode, Codex and other AI tools, the app usually sends instructions and previous conversation history to the model again. If the beginning of the request matches cached data and the model supports caching, that part is billed at the lower Cached input rate instead of regular Input. This is called a cache hit.
The cache is not a saved answer. The model reuses input data but generates a new answer to your request.
- Cache write — creating the cache
Preparing input data for reuse. If the model prices this operation separately, its cost appears in the Cache write column. Cache creation rules and retention depend on the model.
How to configure your client for caching
Choose a suitable API format when your client lets you configure it:
- OpenAI — prefer the Responses API.
- DeepSeek, Qwen, GLM, Kimi and other Chinese models — usually Chat Completions, unless a different format is specified for the selected model.
- Anthropic / Claude — prefer the Messages API, with caching support in your client.
Using a suitable format helps avoid unnecessary request conversions and preserve model features, including caching.
Send a session identifier
If your client supports custom HTTP headers, configure X-Session-Id: use the same value for requests in one conversation and a new one for each new conversation. When supported by the API, this helps associate requests with the same session and use the cache more effectively.
In Hermes Agent, Cline and other clients, this setting depends on the version and selected connection. If the client already sends a session identifier automatically, you do not need to duplicate it.
Model ID — the model name for your connection
Copy the Model ID from the table and use it in the model field of your app or API request.