Ling 3.0 Flash
Ling 3.0 Flash is a 124B Mixture-of-Experts model from Inclusionai activating about 5.1B parameters per token, built for token-efficient agent runs across a context window of 256K tokens.
View API reference- Input and output price
- Input $0.06, Output $0.18, Per 1M tokens
- 24h uptime
- Loading AI Gateway uptime
import { streamText } from 'ai'
const result = streamText({ model: 'inclusionai/ling-3.0-flash', prompt: 'Why is the sky blue?'})Copy link to headingPlayground
Try out Ling 3.0 Flash by Inclusionai. Usage is billed to your team at API rates. Free users (those who haven't made a payment) get $5 of credits every 30 days.
Ling 3.0 Flash
Copy link to headingProviders
Route requests across multiple providers. Copy a provider slug to set your preference. Visit the docs for more info. Using a provider means you agree to their terms, listed under Legal.
Copy link to headingUptime24 hours
Direct request success rate on AI Gateway and per-provider. Visit the docs for more info.
Copy link to headingThroughput24 hours
P50 throughput on live AI Gateway traffic, in tokens per second (TPS). Visit the docs for more info.
Copy link to headingLatency24 hours
P50 time to first token (TTFT) on live AI Gateway traffic, in milliseconds. View the docs for more info.
Getting started
Call Ling 3.0 Flash through AI Gateway with the AI SDK generateText and streamText functions, or through the OpenAI Chat Completions, OpenAI Responses, and Anthropic Messages APIs by changing the base URL. AI Gateway authenticates the request and routes it to an available provider.
Install the AI SDK (pnpm add ai dotenv), create an API key from the API Keys page, and set it as AI_GATEWAY_API_KEY in your environment. Full setup is covered in the text generation quickstart.
import { generateText } from 'ai';import 'dotenv/config';
async function main() { const result = await generateText({ model: 'inclusionai/ling-3.0-flash', prompt: 'Why is the sky blue?', });
console.log(result.text);}
main().catch(console.error);Top-level parameters
The same Ling 3.0 Flash request in each API format AI Gateway supports.
import { generateText } from 'ai';import 'dotenv/config';
async function main() { const result = await generateText({ model: 'inclusionai/ling-3.0-flash', system: 'You are a concise technical assistant.', prompt: 'Summarize the tradeoffs between static generation and SSR.', maxOutputTokens: 1024, });
console.log(result.text);}
main().catch(console.error);Standard parameters like prompt, messages, temperature, and tools work as documented in the AI SDK docs. These are the parameters with model-specific behavior.
| Parameter | Type | Required | Description |
|---|---|---|---|
model | string | Yes | Model ID in the form creator/model, e.g. inclusionai/ling-3.0-flash. AI Gateway routes the request to an available provider. |
maxOutputTokens | number | No | Hard cap on generated tokens. Ling 3.0 Flash supports up to 32,000 output tokens. Reasoning tokens count toward this limit. |
reasoning | 'provider-default' | 'none' | 'minimal' | 'low' | 'medium' | 'high' | 'xhigh' | No | Provider-agnostic reasoning effort, available in AI SDK 7 or later. Maps to the provider’s native reasoning configuration; reasoning settings under providerOptions take precedence when both are set. See the Reasoning section below. |
providerOptions | Record<string, JSONValue> | No | AI Gateway routing options under gateway, plus any provider-native options under the provider’s own namespace — see the table below. |
Input limits
| Input | Formats | Sources | Max count | Max size | Limits |
|---|---|---|---|---|---|
| Text | — | — | — | — | Prompt and response share the 256K-token context window |
Provider options
Set AI Gateway routing options under providerOptions.gateway. For provider-specific options, pass them under the provider’s namespace as documented by the AI SDK.
Learn more in the AI SDK provider docs.
import { generateText } from 'ai';import 'dotenv/config';
async function main() { const result = await generateText({ model: 'inclusionai/ling-3.0-flash', prompt: 'Why is the sky blue?', providerOptions: { gateway: { only: ['novita'], }, }, });
console.log(result.text);}
main().catch(console.error);These AI Gateway routing options apply to every model. Provider-specific options pass through under the provider’s own namespace (for example providerOptions.anthropic) exactly as documented by the AI SDK.
| Parameter | Type | Required | Description |
|---|---|---|---|
providerOptions.gateway.only | string[] | No | Restrict routing to these provider slugs. Requests fail over only within the listed providers. |
providerOptions.gateway.order | string[] | No | Preferred provider order. Listed providers are tried first; unlisted providers remain available as fallbacks. |
providerOptions.gateway.sort | 'cost' | 'ttft' | 'tps' | No | Rank candidate providers by price, time to first token, or tokens per second instead of the default routing order. |
providerOptions.gateway.zeroDataRetention | boolean | No | Route only to providers with a zero-data-retention policy for this model. |
Routing across providers
AI Gateway serves the same model through multiple providers and fails over automatically. order expresses a preference while keeping every provider eligible; only is a hard allowlist — if none of the listed providers are available the request fails instead of falling back.
Options under a provider's own namespace (for example providerOptions.anthropic) are forwarded to that provider with the request. Providers ignore option namespaces that don't apply to them, so it is safe to set provider options alongside gateway routing options.
Reasoning
AI Gateway bridges reasoning across every API format. The AI SDK exposes a provider-agnostic top-level reasoning level (none, minimal, low, medium, high, or xhigh); the Chat Completions and Responses formats take the same effort under reasoning.effort; and the Anthropic Messages format uses a native thinking token budget. Whichever you send, the gateway maps it to the target model’s native configuration, converting between effort levels and token budgets as needed. Reasoning-related settings under providerOptions take full precedence over the top-level reasoning value and are never merged. Reasoning tokens typically count toward your output-token usage, though how they’re reported and billed varies by provider.
Learn more in the AI Gateway reasoning guide.
import { generateText } from 'ai';import 'dotenv/config';
async function main() { const result = await generateText({ model: 'inclusionai/ling-3.0-flash', prompt: 'Explain the Monty Hall problem step by step.', reasoning: 'high', });
console.log(result.text);}
main().catch(console.error);Tool calling
Expose tools the model can call. Define each tool’s inputs with a Zod schema.
import { generateText, tool } from 'ai';import { z } from 'zod';import 'dotenv/config';
async function main() { const result = await generateText({ model: 'inclusionai/ling-3.0-flash', prompt: 'What is the weather in San Francisco?', tools: { getWeather: tool({ description: 'Get the current weather for a location', inputSchema: z.object({ location: z.string() }), execute: async ({ location }) => ({ location, temperatureC: 18 }), }), }, });
console.log(result.text);}
main().catch(console.error);Copy link to headingMore models by Inclusionai
| Model |
|---|
| 256K | 2.6 s | 201 tps | Free | Free | — |  | 09/04/2026 | |||||||
| 262K | 0.3 s | 406 tps | Free | Free | Free | — |  | 08/27/2026 |
Copy link to headingAbout Ling 3.0 Flash
Ling 3.0 Flash is Inclusionai's hybrid reasoning model, built around sparsity: 124 billion total parameters with about 5.1 billion active per token, roughly a one-in-sixty-four expert activation. Inclusionai positions it as matching or beating its own much larger flagship on most benchmarks while activating a fraction of the parameters.
The attention design is hybrid from the start rather than retrofitted. A repeating five-to-one stack alternates Kimi Delta Attention with multi-head latent attention, so linear attention keeps long inputs cheap while periodic full-attention layers preserve exact token-to-token recall. That combination is what lets a model this sparse hold up on reasoning tasks instead of only on throughput.
The context window is 256K tokens, with up to 32K tokens per response. Thinking mode is enabled by default and the model scales its thinking effort to the difficulty of the task, so simple prompts do not pay the full reasoning cost.
Ling 3.0 Flash scores 38 on the Artificial Analysis Intelligence Index, well above the median for open-weight models of comparable size. Weights are published on Hugging Face.
You can integrate Ling 3.0 Flash through AI SDK, Chat Completions API, Responses API, Messages API, or other API formats, from TypeScript or Python.
Copy link to headingWhat To Consider When Choosing a Provider
- Configuration: Inclusionai's launch comparison against a one-trillion-parameter flagship drew scrutiny. The comparison point appears to be a specific expert variant rather than the headline 1T model the launch text implied, and the launch chart was published without a readable data table. Treat the parameter-efficiency claim as directionally real and the precise margin as unverified.
- Configuration: Thinking is on by default. That is usually what you want on agent work, but it means a trivial prompt still spends reasoning tokens unless you turn it off, so check your output token budget on high-volume simple calls.
- Configuration: There is also a free routing tier,
ling-3.0-flash-free, serving the same model. Compare the two on this page before you commit paid traffic, since for evaluation and prototyping the free tier is the cheaper path to the same output. - Zero Data Retention: Zero Data Retention is offered on a per-provider and model basis. See the documentation for details.
- Authentication: AI Gateway authenticates requests using an API key or OIDC token. You do not need to manage provider credentials directly.
Copy link to headingWhen to Use Ling 3.0 Flash
Best for
- Token-Efficient Agent Runs: About 5.1B active parameters keeping long loops affordable
- Long-Context Reasoning: Hybrid attention that stays cheap without losing exact recall
- Open-Weight Deployments: Published weights on Hugging Face
- Difficulty-Scaled Thinking: Reasoning effort proportional to the task
- Cost-Sensitive Throughput: Sparsity beating a dense model of similar capability
Consider alternatives when
- Evaluation And Prototyping:
ling-3.0-flash-freeserves the same model at no cost - High-Volume Simple Calls: Thinking is on by default and spends tokens
- Frontier-Tier Reasoning: Its Artificial Analysis score sits below the leaders
- Verified Comparisons: The launch head-to-head claim was not fully substantiated
Copy link to headingConclusion
Ling 3.0 Flash is a sparse Mixture-of-Experts model that keeps long agent runs cheap, activating about 5.1B of 124B parameters per token across a 256K tokens window. Point inclusionai/ling-3.0-flash at AI Gateway for paid traffic, and use ling-3.0-flash-free while you are still evaluating.
Copy link to headingFrequently Asked Questions
What is Ling 3.0 Flash built for?
Token-efficient agentic inference. About 5.1 billion of its 124 billion parameters activate per token, so multi-step runs fit inside tighter cost and latency budgets.
How is Ling 3.0 Flash different from ling-3.0-flash-free?
It is the same model on a paid routing tier. Use the free tier for evaluation and prototyping, and this one for production traffic. Compare both on this page.
What is the context window for Ling 3.0 Flash?
The context window is 256K tokens, with up to 32K tokens per response.
How does the hybrid attention design work?
A repeating five-to-one stack alternates Kimi Delta Attention with multi-head latent attention. Linear attention keeps long inputs cheap, and periodic full-attention layers preserve exact token-to-token recall.
Is thinking mode on by default?
Yes, and the model scales thinking effort to task difficulty. On high-volume simple calls, check your output token budget, since reasoning tokens are still spent unless you disable it.
Are the weights open?
Yes. Inclusionai publishes the weights on Hugging Face.
Does Ling 3.0 Flash really match a one-trillion-parameter model?
The parameter-efficiency claim is directionally supported but the precise margin is not. The launch comparison point appears to be a specific expert variant rather than the headline flagship, and no readable data table accompanied the chart.
Does Ling 3.0 Flash support Zero Data Retention?
Zero Data Retention is not currently available for this model. Zero Data Retention is offered on a per-provider basis. See https://vercel.com/docs/ai-gateway/capabilities/zdr for details.