Ling 3.0 Flash
Ling 3.0 Flash is a Mixture-of-Experts model from Inclusionai with 124B total parameters and about 5.1B active per token, built for token-efficient agentic inference. It supports a context window of 256K tokens and a max output of 256K tokens per request.
import { streamText } from 'ai'
const result = streamText({ model: 'inclusionai/ling-3.0-flash-free', 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.
Qwen 3 32B
Qwen 3 32B
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 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.
Copy link to headingUptime24 hours
Direct request success rate on AI Gateway and per-provider. Visit the docs for more info.
Copy link to headingMore models by Inclusionai
| Model |
|---|
| 256K | 1.0s | 419tps | Free | Free | — |  | 08/06/2026 | ||||||
| 256K | 0.7s | 310tps | $0.06/M | $0.18/M | Read:$0.01/M Write:— | — | | 08/06/2026 |
Copy link to headingAbout Ling 3.0 Flash
Ling 3.0 Flash arrived on AI Gateway on July 23, 2026. The architecture is a Mixture-of-Experts (MoE) with 124B total parameters and about 5.1B active per token, so inference cost tracks the active subset rather than the full parameter count. The context window is 256K tokens, and a single request can return up to 256K tokens.
Inclusionai built Ling 3.0 Flash around token efficiency and production-scale agentic inference, meaning more completed work per token spent across long multi-step runs. The target workloads are high-frequency agentic workflows, coding agents, document processing, and long-context multi-turn conversation. Ling 3.0 Flash runs in thinking and non-thinking modes, so you spend reasoning tokens on the requests that need them.
The -free suffix in the model identifier inclusionai/ling-3.0-flash-free marks a free listing. Ling 3.0 Flash joined the catalog at no cost for three weeks after launch, through August 3, 2026. AI Gateway currently lists $0 per million input tokens and $0 per million output tokens, and the pricing panel on this page always shows the live rate. Inclusionai later shipped Ling 3.0 Tiny, a smaller sibling with 7.9B total parameters and about 1.3B active per token, which took over the free slot on August 6, 2026.
Call Ling 3.0 Flash with one API key and get provider routing, automatic failover, and built-in observability. Integrate through the AI SDK, the Chat Completions API, the Responses API, the Messages API, or other supported API formats. AI Gateway reflects provider pricing with no markup and charges no platform fee on inference.
Copy link to headingWhat To Consider When Choosing a Provider
- Configuration: Ling 3.0 Flash handles text input and text output. Route requests that carry images, audio, or video to a multimodal model instead.
- Configuration: Token efficiency is Inclusionai's stated design goal rather than a measured guarantee for your workload, so run your own agent traces before you move production traffic. For current throughput and latency, see live metrics on this page.
- Configuration: The
-freesuffix is part of the identifier you send, not a flag you set. Pin the identifier you intend to keep so your calls resolve to the listing you expect. - Zero Data Retention: AI Gateway does not currently support Zero Data Retention for this model. See the documentation for models that support ZDR.
- 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
- High-Frequency Agent Loops: Repeated small steps that must stay inside a fixed token budget
- Cost-Sensitive Coding Agents: Repository work where a small active footprint holds per-request cost down
- Long Multi-Turn Sessions: Extended conversation that uses the full 256K tokens window
- Document Processing: Long reports, transcripts, and mixed-format files that need a wide input window
- Mixed Reasoning Traffic: Workloads that pair quick replies with occasional deep reasoning
Consider alternatives when
- Multimodal Inputs: Ling 3.0 Flash handles text only, so send images, audio, or video to a vision-capable model
- Benchmark-Driven Selection: No published scores accompany this listing, so plan your own evaluation
- Frontier Capability Ceilings: Larger flagship models still lead on the hardest coding and research suites
- Stable Model Identifiers: The
-freesuffix marks a free listing, so pin the identifier you intend to keep
Copy link to headingConclusion
Ling 3.0 Flash gives you a sparse MoE agent model with a context window of 256K tokens and two reasoning modes, tuned for agent work that repeats at high volume. Route it through AI Gateway and you get failover, observability, and one key across the whole catalog.