Sulat.com
AI models
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3 from Synthetic
LLM Gateway logo

Model details

Qwen3 Next 80B A3B Thinking (Vertex AI (OpenAI-compatible))

Qwen3 Next 80B A3B Thinking is positioned as an open-weight large language model in Alibaba's Qwen family, designed around explicit reasoning behavior and tool use rather than general chat alone. Third-party cataloging groups it under chat, long-context, reasoning, tool-use, and open-weight categories, reflecting an intended role as a thinking-style assistant that can plan through multi-step problems and call external functions or APIs when needed. The "Thinking" suffix in the variant name signals a reasoning-oriented mode of operation, distinguishing it from standard chat-tuned Qwen releases and aligning it with developer workflows that benefit from visible chain-of-thought style traces before final answers.

In practical terms, the model is aimed at workloads that combine extended document understanding with structured tool use, such as analyzing long reports, orchestrating agentic pipelines, and answering questions that require pulling in external data through function calls. Its open-weight status makes it attractive to teams that want to self-host, fine-tune, or audit the model rather than rely solely on a remote API, while still working within the Qwen ecosystem. Because its reasoning and tool-use orientation is documented alongside a long-context profile, it fits well as a backbone for assistants that need to reason over large inputs and then act on the conclusions, rather than as a lightweight conversational model for short, casual exchanges.

LLM Gatewayvertex-openai/qwen3-next-80b-a3b-thinkingqwen

Quick Info

Powered by
Provider
LLM Gateway
Model key
vertex-openai/qwen3-next-80b-a3b-thinking
Release date
Sep 1, 2025
Last updated
Sep 1, 2025
Knowledge cutoff
2025-04
Input modalities
Output modalities
Capabilities

Cost

Input token cost
$0.15
Output token cost
$1.20

Limits

Output tokens
32,768 tokens
Context window
131,072 tokens

Transparent token rates

Compare Qwen3 Next 80B A3B Thinking (Vertex AI (OpenAI-compatible)) pricing

Rates are shown per one million tokens. Combined means one million input plus one million output tokens.

Browse this family

Latest news about Qwen3 Next 80B A3B Thinking (Vertex AI (OpenAI-compatible))

LLM Gateway

CoverageBenchmark

An NVIDIA Developer Blog post (dated September 15, 2025) documents the open-source release of Alibaba's Qwen3-Next-80B-A3B-Thinking model alongside its Instruct sibling, previewing a hybrid Mixture of Experts architecture. The Thinking variant is explicitly named, and the post positions it as a next-generation reasonin Technically, the blog specifies that the model has 80B total parameters but activates only about 3B per token, using 512 routed experts plus 1 shared expert across a 48-layer design that alternates Gated Delta Network / GQA attention every fourth layer with linear attention in the remaining layers. It supports context

Videos about Qwen3 Next 80B A3B Thinking (Vertex AI (OpenAI-compatible))

More models around Qwen3 Next 80B A3B Thinking (Vertex AI (OpenAI-compatible))