---
title: "Unlimited LLM API for agents - sessions explained | BroadRouter"
description: "The unlimited DeepSeek V4.1 Flash plans have no token meter and no weekly limit. Here is how sessions, parallel agents, keys and billing actually work."
url: https://broadrouter.com/guides/unlimited-llm-api-sessions/
lang: en
updated: 2026-11-16
publisher: BroadRouter (a BroadNet site); API by BroadNet Inference, https://inference.broadnet.ai
---

# How the unlimited LLM API plans work for agents

Updated 2026-11-16

Each plan session gives unlimited DeepSeek V4.1 Flash tokens, one request at a time. There is no token meter, no monthly cap, no 5-hour window and no weekly limit. To run agents in parallel, buy one session per agent. Plans last a month, a week or an hour.

BroadNet’s plans have no token meter, no rolling 5-hour window and no weekly quota. They cap only one thing: how many requests run **at the same time**.

## The three plans

| Plan | Price per session | Lasts | Paid by |
| --- | --- | --- | --- |
| Full Moon | $60 | a month, renews | Card |
| Orbit | $15 | a week; buy the next week before it ends to keep the same key | Account balance |
| Meteor | $1 | an hour from purchase | Account balance |

All three are the same model, DeepSeek V4.1 Flash, with a 1M-token context and typical speed of about 100 tokens per second. With a founders code, Full Moon is $45 a month for the first 3 months. [Compare a plan with pay-as-you-go](https://broadrouter.com/pricing/).

## What a session is

A session runs one request at a time, the way one agent works: it asks, reads the answer, and asks again. Plan for one request in flight per session.

-   One coding agent in one terminal: **one session**.
-   An orchestrator plus three sub-agents running at once: **four sessions**.
-   Ten batch workers in parallel: **ten sessions**.

Every session costs the same, and one key can hold several. Give each parallel agent its own session.

## Keys and plans

Each plan has its own keys, and you split its sessions across them. Requests on a plan key never draw on your balance. Your pay-as-you-go keys and [hourly keys, explained on the pricing page](https://broadrouter.com/pricing/), keep working separately. Your $5 starting credit covers pay-as-you-go and hourly keys; plans are bought separately.

## Try it before a month

Buy a Meteor hour ($1) and run your real agent loop on it. Point the agent at `https://inference.broadnet.ai/v1` with model `broadnet/deepseek-v4.1-flash`; for a terminal coding agent, see [the Aider with unlimited DeepSeek guide](https://broadrouter.com/guides/aider-unlimited-deepseek/).

## How does this compare with other flat-rate LLM APIs?

Several providers sell monthly plans. They limit different things, so compare what you will actually run. Prices below were read from each provider’s pricing page on 2026-11-14.

| Provider and plan | Price | What is limited | Agents and API traffic allowed? |
| --- | --- | --- | --- |
| BroadNet Full Moon | $60 a month per session | One request in flight per session; one model (DeepSeek V4.1 Flash) | Yes, for your own agents and products |
| Featherless Chat | USD 25 a month | Unlimited tokens, 32K context, 4 concurrent units | No: interactive chat only, no app or API traffic or background automation |
| Featherless Developer | from USD 50 a month in credits | Billed per token, context up to 256K | Yes, but it is metered, not flat |
| Awan LLM Pro | USD 20 a month | Unlimited tokens; 100 requests a minute and daily request caps per model size | Yes |
| Awan LLM Max | USD 80 a month | Unlimited tokens; 200 requests a minute, no daily caps | Yes |
| Standard Compute Starter / Economy | USD 19 / USD 39 a month | A compute budget equal to the price; requests stop when it runs out | Yes; Starter runs one agent at a time |

Where they beat BroadNet: Awan LLM Pro is cheaper per month and allows parallel requests on one plan, and Featherless offers a catalogue of 40,000+ open models instead of one. Standard Compute routes to frontier models such as Claude and GPT, which BroadNet does not serve. Where BroadNet differs: no daily request cap and no token budget on a session, and the plan explicitly covers coding agents. Awan LLM’s pricing page footer is dated 2023-2024 and it lists no context size there, so check its models page before you buy.

## Troubleshooting

-   **Requests wait or fail when agents run in parallel.** One session runs one request at a time. Count the agents that call the model at the same moment, including sub-agents, and buy that many sessions.
-   **Empty answers.** DeepSeek V4.1 Flash reasons first, and reasoning counts toward `max_tokens`. Leave room or set `reasoning_effort` to `low`. See [why reasoning models return empty answers](https://broadrouter.com/guides/reasoning-tokens-empty-response/).
-   **A plan key fails on another model.** Plans cover `broadnet/deepseek-v4.1-flash` only. Use a pay-as-you-go key for other models.
-   **The Meteor hour ended mid-task.** Meteor starts at purchase and lasts one hour. For longer runs use Orbit or Full Moon.
-   **Errors or limits you don’t recognise.** Check the status code against [BroadNet error codes and retries](https://broadrouter.com/guides/errors-and-retries/).

## What it is not for

The plan is for your own agents and the products you build. Sharing or reselling the key isn’t allowed, and access follows BroadNet’s [Acceptable Use Policy](https://www.broadnet.ai/aup).

[Choose a plan on BroadNet Inference](https://inference.broadnet.ai/plans?utm_source=broadrouter&utm_medium=referral&utm_campaign=guides-unlimited-llm-api-sessions&utm_content=body)

## Terms in this guide

[Token](https://broadrouter.com/glossary/token/)

A token is the unit of text a language model reads and writes: a word, part of a word, a punctuation mark or a byte sequence produced by the model's tokenizer. LLM APIs count, limit and usually bill usage in tokens.

[Pay-as-you-go](https://broadrouter.com/glossary/pay-as-you-go/)

Pay-as-you-go is a pricing model in which you pay only for what you use, metered per unit such as tokens, requests or audio minutes, with no subscription or minimum commitment. Most LLM APIs bill this way, from a prepaid balance or a periodic invoice.

[AI coding agent](https://broadrouter.com/glossary/coding-agent/)

An AI coding agent is a software tool in which a language model works on a codebase in a loop: it reads files, edits them, runs commands such as tests, reads the output and decides the next step until the task is done or a limit is reached.

[Prepaid credit](https://broadrouter.com/glossary/prepaid-credit/)

Prepaid credit is money added to an API account before use. Each request deducts its cost from the balance, and when the balance runs out, metered requests stop until you top up. It caps spend at what you have paid in.

[Subagent](https://broadrouter.com/glossary/subagent/)

A subagent is a separate agent that a main agent starts to handle one focused task. It runs with its own context window, system prompt and tool access, does the work, and returns only a result or summary to the main agent, which keeps the main conversation small.

[AI agent](https://broadrouter.com/glossary/agent/)

An AI agent is a system in which a language model directs its own steps toward a goal: it decides which tools to call, reads the results and repeats until it produces a final answer or hits a limit. A fixed script that calls a model at set points is usually called a workflow instead.

[Concurrency](https://broadrouter.com/glossary/concurrency/)

Concurrency is the number of requests a client has in flight at the same time. LLM APIs often cap it per key or plan, separately from requests per minute, because each open request holds GPU capacity until its response has finished.

[Context window](https://broadrouter.com/glossary/context-window/)

A context window is the maximum number of tokens a language model can work with in a single request. It covers the prompt, conversation history, tool definitions, images and the generated output, including any reasoning tokens. Anything that does not fit must be cut, summarised or split across several requests.

## Questions

### Is there a cheaper unlimited LLM API than BroadNet?

Yes, on sticker price. On 2026-11-14 Awan LLM listed a Pro plan at USD 20 a month with unlimited tokens but per-minute and daily request caps, and Featherless listed a USD 25 chat plan that bans agent and API traffic. Compare the limits, not only the price.

### Can I share a plan key with my team?

No. Each person uses their own account. You can build the plan into your own app, agent or service, but you cannot share or sell the key.

### Do plan requests use my balance?

No. Requests on a plan key never draw on your balance, and your pay-as-you-go keys keep working as before.

### Can I cancel?

Monthly and weekly renewal is managed in the console. Paid access continues to the end of the purchased period. Statutory refund rights apply.

### Is there really no token limit on the unlimited plan?

BroadNet's plans page says DeepSeek V4.1 Flash has no token meter and no monthly cap on any plan, with no 5-hour window and no weekly limit. The limit is concurrency, one request at a time per session.

### Does the plan work with coding agents like Aider, Cline or OpenCode?

Any tool that accepts an OpenAI-compatible base URL and a custom model ID can use it. Point the tool at https://inference.broadnet.ai/v1 with model broadnet/deepseek-v4.1-flash and your plan key.

## Related guides

-   [Run Aider on unlimited DeepSeek V4.1 Flash](https://broadrouter.com/guides/aider-unlimited-deepseek/)
-   [Why a reasoning model returns an empty answer](https://broadrouter.com/guides/reasoning-tokens-empty-response/)
-   [Use the OpenAI Python SDK with BroadNet Inference](https://broadrouter.com/guides/openai-python-sdk/)

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=guides-unlimited-llm-api-sessions&utm_content=page-cta) ·

**Sources**

-   [BroadNet Inference unlimited plans and plan FAQ](https://inference.broadnet.ai/plans?utm_source=broadrouter&utm_medium=referral&utm_campaign=guides-unlimited-llm-api-sessions&utm_content=sources)
-   [Machine-readable plan prices (pricing.json)](https://inference.broadnet.ai/pricing.json?utm_source=broadrouter&utm_medium=referral&utm_campaign=guides-unlimited-llm-api-sessions&utm_content=sources)
-   [Featherless pricing (checked 2026-11-14)](https://featherless.ai/pricing)
-   [Awan LLM pricing (checked 2026-11-14)](https://www.awanllm.com/pricing)
-   [Standard Compute pricing (checked 2026-11-14)](https://standardcompute.com/pricing)
