---
title: "What are reasoning tokens? Billing, limits, empty answers"
description: "Reasoning tokens are what a reasoning model writes while thinking before it answers. How they are billed, how they count toward max_tokens, and how to control them."
url: https://broadrouter.com/glossary/reasoning-tokens/
lang: en
updated: 2026-11-14
publisher: BroadRouter (a BroadNet site); API by BroadNet Inference, https://inference.broadnet.ai
---

# Reasoning tokens

Updated 2026-11-14

Reasoning tokens are tokens a reasoning model generates to work through a problem before it writes its visible answer. They are usually billed as output tokens, count toward the output limit and the context window, and may be hidden or summarized by the API.

Also called: thinking tokens, chain-of-thought tokens, reasoning effort, extended thinking

Reasoning models think before they answer. That thinking is text too, so it costs tokens, time and space in the context window.

## How it works

A reasoning model generates an internal chain of thought, then the visible answer. Both come from the same output budget. OpenAI’s guide says reasoning tokens “are billed as output tokens” and occupy space in the context window; Anthropic says the same of thinking tokens and calls them “a subset of your `max_tokens`.”

In the Chat Completions format, the count appears in usage:

```json
"usage": {
  "completion_tokens": 1186,
  "completion_tokens_details": { "reasoning_tokens": 1024 }
}
```

Here, 1,024 of 1,186 output tokens went to thinking, and only 162 to the answer. OpenAI warns that a response can hit the limit “before any visible output tokens are produced,” so you pay for input and reasoning and get no reply. It suggests reserving at least 25,000 tokens for reasoning and output when you start experimenting with its models.

Many APIs expose a control such as `reasoning_effort` (`low`, `medium`, `high`) to trade quality for fewer thinking tokens.

## Why it matters

-   **Cost.** On a per-token key, reasoning is billed at the output price, usually the highest rate. See [input and output tokens](https://broadrouter.com/glossary/input-and-output-tokens/).
-   **Latency.** Thinking takes time before the first visible token. At a fixed [tokens per second](https://broadrouter.com/glossary/tokens-per-second/), 2,000 reasoning tokens add the time of 2,000 output tokens.
-   **Reliability.** A small [`max_tokens`](https://broadrouter.com/glossary/max-tokens/) that worked on a non-reasoning model can return empty answers on a reasoning one.

## On BroadNet Inference

[DeepSeek V4.1 Flash](https://broadrouter.com/models/deepseek-v4.1-flash/), [GLM 5.3](https://broadrouter.com/models/glm-5.3/) and [Qwen3.8 27B](https://broadrouter.com/models/qwen3.8-27b/) reason before they answer. BroadNet’s docs say the reasoning counts toward `max_tokens`. BroadNet’s docs do not list reasoning for BroadNet Gemma 4 VL 26B.

When the model reports it, `usage.completion_tokens_details.reasoning_tokens` shows the thinking share. Set `max_tokens` to 1,000 or more, or leave it unset, and use `reasoning_effort="low"` for simple steps:

```python
client.chat.completions.create(
    model="broadnet/deepseek-v4.1-flash",
    reasoning_effort="low",
    max_tokens=1000,
    messages=[{"role": "user", "content": "Classify: 'refund my order'"}],
)
```

On pay-as-you-go, reasoning is billed at the output price, for example $0.52 per 1M tokens on DeepSeek V4.1 Flash. On an unlimited plan for that model there is no token meter, so reasoning costs nothing extra; it only takes time. The full fix list is in [why a reasoning model returns an empty answer](https://broadrouter.com/guides/reasoning-tokens-empty-response/).

## Questions

### Can I see the reasoning tokens?

It depends on the API. OpenAI says reasoning tokens are not visible through its API but can be summarized. Other providers return some or all of the reasoning in a separate field. You are billed for them either way.

### How do I reduce reasoning tokens?

Use a lower reasoning effort setting for simple tasks, such as reasoning\_effort set to low, where the model supports it. Clear, specific prompts also help.

### Why did I get an empty answer from a reasoning model?

The model probably spent the whole output budget on reasoning. finish\_reason is then length. Raise max\_tokens or lower the reasoning effort.

## See also

-   [max\_tokens](https://broadrouter.com/glossary/max-tokens/) (term)
-   [Input and output tokens](https://broadrouter.com/glossary/input-and-output-tokens/) (term)
-   [Context window](https://broadrouter.com/glossary/context-window/) (term)
-   [Tokens per second](https://broadrouter.com/glossary/tokens-per-second/) (term)
-   [Chain-of-thought (CoT)](https://broadrouter.com/glossary/chain-of-thought/) (term)
-   [Why a reasoning model returns an empty answer](https://broadrouter.com/guides/reasoning-tokens-empty-response/) (guide)
-   [Stream chat completions and handle dropped connections](https://broadrouter.com/guides/streaming-chat-completions/) (guide)
-   [Switching to an OpenAI-compatible API: a migration checklist](https://broadrouter.com/blog/openai-compatible-api-migration-checklist/) (article)
-   [Flat-rate vs pay-per-token LLM API: the break-even math](https://broadrouter.com/blog/flat-rate-vs-pay-per-token-llm-api/) (article)

Try it on your real workload. New accounts start with $5 of credit for Gemma and transcription. DeepSeek, GLM and Qwen3.8 need a top-up.

[Get started](https://inference.broadnet.ai/account/?utm_source=broadrouter&utm_medium=referral&utm_campaign=glossary-reasoning-tokens&utm_content=page-cta) ·

**Sources**

-   [OpenAI, reasoning models guide](https://developers.openai.com/api/docs/guides/reasoning)
-   [Anthropic, context windows with thinking](https://platform.claude.com/docs/en/build-with-claude/context-windows)
-   [BroadNet Inference API reference, reasoning and cached tokens](https://inference.broadnet.ai/docs?utm_source=broadrouter&utm_medium=referral&utm_campaign=glossary-reasoning-tokens&utm_content=sources#models)
