O3 API

OpenAI's O3, through an OpenAI-compatible API. Providers quote it on an open exchange, and each request goes to the offer with the lowest estimated charge that meets your rules.

Providers compete for each request. The exchange ranks the offers that meet your rules by estimated charge. It fixes your maximum charge before the first token, and you pay for the tokens that the model uses.
The OpenAI and Anthropic APIs. Use the OpenAI or Anthropic SDK that you have. Set the base URL and your key, and send o3 as the model.
Your data rules. Send x-liquid-retention: none and only offers whose provider does not keep your content compete for the request.

Facts

Model name: o3
Maker: OpenAI
Context window: 200K tokens
Input price: $1.997 per million tokens
Output price: $7.987 per million tokens
Providers: 1 independent provider
Regions: us

Providers

Architect: us, $1.997 input and $7.987 output per million tokens, $0.998 cache read per million tokens, 200K tokens of context

Scores

HLE: 20.1, #117 of 273
GPQA: 82.7, #90 of 250
AA-LCR: 74.7, #90 of 237
Terminal-Bench Hard: 37.1, #33 of 166
LiveCodeBench: 80.8, #19 of 130
AIME 2025: 88.3, #15 of 100
MMLU-Pro: 85.3, #15 of 133
τ²-Bench: 80.7, #63 of 169
IFBench: 71.4, #35 of 178

Scores: Artificial Analysis (https://artificialanalysis.ai/)

Request

curl https://router.inference.ai.exchange/v1/chat/completions \
  -H "Authorization: Bearer $LIQUID_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "o3",
    "messages": [{"role":"user","content":"Say hello."}]
  }'

Questions

What is O3?: Served by intent-router: each job goes to the cheapest healthy upstream for this model.
How much does O3 cost?: Offers now ask $1.997 per million input tokens and $7.987 per million output tokens. A job pays for the tokens it uses. The exchange picks the cheapest eligible offer.
What is the context length of O3?: The largest offer takes 200,000 input tokens. A smaller offer can take less. The exchange skips an offer that cannot take your message.
What can O3 do?: Its providers list Reasoning effort, JSON object, JSON schema, Frequency penalty, Logit bias, Log probabilities, Min-p sampling, Presence penalty, Repetition penalty, Seed, Stop sequences, Top-k sampling, Top-p sampling, Text, Tools.
What inputs and outputs does O3 support?: Text → Text.
Which providers serve O3?: 1 independent provider quote it now, with 1 offer. The table above ranks them by the cost of a typical job.
How do I call O3?: Send "o3" as the model in an OpenAI or Anthropic request. The API section links to setup steps for each client.
When was O3 listed?: The exchange first listed it on 10/8/2026. This is not its release date.