LLM Gateway

Models and pricing

Every model available through the LLM Gateway, with context windows and per-token pricing.

View as Markdown

The gateway routes one model catalog across OpenAI, Anthropic, xAI, and open-weight providers. Pass any ID below as model. An unknown ID returns 400.

Prices are USD per million tokens. Cache write and cache read apply to prompt caching; models without cache-write pricing still discount cached reads.

Query the live catalog at any time:

curl https://llm.clusterbase.dev/v1/models \
  -H "Authorization: Bearer $CLUSTER_API_KEY"

Each entry includes the context window, default output limit, supported reasoning levels, supported speeds, and the standard rate card.

Current models

ModelContextInputOutputCache writeCache read
gpt-6-astra1,050,000$10.00$50.00$12.50$1.00
gpt-6.1-sol1,050,000$2.00$10.00$2.50$0.10
gpt-6-luna1,050,000$0.10$0.50$0.125$0.01
claude-fable-5-11,000,000$10.00$50.00$12.50$0.25
claude-opus-5-51,000,000$4.00$20.00$5.00$0.20
claude-sonnet-5-51,000,000$2.00$10.00$2.50$0.20
claude-haiku-4-5200,000$1.00$5.00$1.25$0.10
kimi-k31,048,576$3.00$15.00—$0.30
deepseek-v4.1-flash1,048,576$0.22$0.66—$0.007
qwen3p8-max262,144$2.00$6.00—$0.25
grok-4.7500,000$2.00$6.00—$0.50

Aliases

Two product aliases track our recommended defaults, so you get model upgrades without a code change:

AliasCurrently routes to
chat-model-smartgpt-6.1-sol
chat-model-fastgpt-6-luna

Aliases run without reasoning; a request that sets reasoning_effort on an alias returns 400. Use a concrete model ID when you need reasoning control.

Fast mode

GPT-6 Astra, GPT-6.1 Sol, GPT-6 Luna, and kimi-k3 accept "speed": "fast" for lower-latency inference. Fast requests bill at twice the standard rate (1.5× for kimi-k3). All other current models run at standard speed only; check supported_speeds in /v1/models.

Long-context pricing

For GPT-6 Astra, GPT-6.1 Sol, and GPT-6 Luna, a request whose input exceeds 272,000 tokens bills the whole request at multiplied rates: input at 2× and output at 1.5×.

For grok-4.7, a request whose input reaches 200,000 tokens bills the whole request at 2× input, cached input, and output.

Reasoning

Most models accept reasoning_effort (for example low, medium, high). The accepted levels differ per model — read reasoning_levels from /v1/models. claude-haiku-4-5 does not support reasoning. Reasoning output is only emitted on streaming requests.

GPT-6 Astra (gpt-6-astra) and GPT-6.1 Sol (gpt-6.1-sol) accept low, medium, high, xhigh, and max; omitted effort defaults to medium. They do not support temperature or top_p — requests that set either are rejected with 400.

GPT-6 Luna (gpt-6-luna) accepts none, low, medium, high, xhigh, and max; omitted effort defaults to medium. It accepts temperature and top_p only with "reasoning_effort": "none".

claude-fable-5-1, claude-opus-5-5, and claude-sonnet-5-5 also accept low, medium, high, xhigh, and max, with adaptive thinking always on; omitted effort defaults to medium for claude-opus-5-5 and high for the other two. They reject requests that set both temperature and top_p, force tool_choice, or end with an assistant message (prefill).

Cluster Bot (Computer) sessions admit only gpt-6.1-sol and gpt-6-luna; every other model, including the aliases, is rejected with 400.

grok-4.7 accepts low, medium, high, and xhigh; omitted effort defaults to high. It accepts image attachments.

Deprecated models

These IDs still resolve but are scheduled for removal. Migrate to a current model.

ModelInputOutputReplacement
gpt-5.6$4.00$20.00gpt-6.1-sol
gpt-5.6-sol$4.00$20.00gpt-6.1-sol
gpt-5.6-terra$2.00$12.00gpt-6.1-sol
gpt-5.6-luna$0.20$1.20gpt-6-luna
gpt-5.5$5.00$30.00gpt-6.1-sol
gpt-5.4$2.50$15.00gpt-6.1-sol
gpt-5.2$1.75$14.00gpt-6-luna
claude-opus-5$5.00$25.00claude-opus-5-5
claude-opus-4-8$5.00$25.00claude-opus-5-5
claude-opus-4-7$5.00$25.00claude-opus-5-5
claude-sonnet-5$2.00$10.00claude-sonnet-5-5
claude-sonnet-4-6$3.00$15.00claude-sonnet-5-5
claude-fable-5$10.00$50.00claude-fable-5-1

/v1/models marks these with "deprecated": true.

On this page