Qwen Plus Pricing (2026)

Qwen-Plus is Alibaba's mid-tier general model, and measured via OpenRouter it routes at $0.26 input / $0.78 output per million tokens, roughly a third of the Qwen3-Max routed rate while clearing most everyday tasks. That gap is the part most buyers miss: for chat, summarization, and routine extraction, Plus often wins on cost-per-acceptable-answer, and you only step up to Max when a task actually fails on Plus. Below are the routed numbers, the official Alibaba rates, and one live call we ran ourselves.

Qwen-Plus pricing (verified 2026-06)

Qwen-Plus is the workhorse tier that sits between the cheap small models and the flagship Max. The price you pay depends entirely on where you buy it, which is where most comparison tables go wrong. There are two distinct numbers, and blurring them is how buyers overpay.

The first is the OpenRouter-routed price. Measured against the live OpenRouter catalog on 2026-06-26, Qwen-Plus routes at $0.26 per million input tokens and $0.78 per million output. This is the number a developer reaching Qwen through an aggregator actually pays, and it is the basis for every first-hand figure in this article.

The second is Alibaba's own published rate. According to Alibaba Cloud Model Studio pricing, Qwen-Plus on the official endpoint runs roughly $0.40 input and $1.20 output per million tokens. We have not yet re-verified that figure on a native DashScope key in this environment, so treat it as sourced-but-pending, not confirmed first-hand.

The routed Plus rate is materially below the official Alibaba rate, which is unusual but not unique to Plus. For the flagship Max, the gap is even wider in the buyer's favor, so the aggregator route is worth pricing before you assume the first-party endpoint is cheapest.

US and international developers can still reach Qwen on the first-party side if they prefer it. According to Alibaba Cloud Model Studio pricing, Qwen is served through Alibaba Cloud Model Studio International, including a Singapore endpoint with a free quota of about one million tokens for ninety days and a US Global deployment without that free quota. The free quota only softens the official rate temporarily, so for steady production volume the routed price is the one to budget against.

We called Qwen-Plus on the live API (first-hand evidence)

Published rates tell you the headline. They do not tell you what a real request costs once a model decides how verbose to be, which is the number that actually lands on your invoice. So we sent a short general-purpose prompt to Qwen-Plus and measured the billed cost and latency directly.

The call was routed through OpenRouter, which served it as qwen/qwen-plus-2025-01-25. We disclose the routing because that is the price basis; this is not Alibaba's official endpoint.

The single call billed $0.0000832 for 38 input and 94 output tokens, returning in 2.47 seconds. What stands out is the shape of the bill: output tokens did most of the work, which is exactly why the $0.78 output rate matters more than the $0.26 input rate for any generative workload. A model that answers concisely on Plus is cheaper twice over, once on the rate and once on the token count.

For contrast, the same prompt style sent to Qwen3-Max billed $0.00027534 for 38 in and 63 out at 3.22 seconds in our run. Max produced a shorter completion yet cost over three times as much, because its routed output rate is five times higher. That is the cost-per-quality crossover in one data point: unless Max's answer is meaningfully better for your task, Plus delivers the same job for a fraction of the spend.

When Plus beats Max on cost-per-quality

The decision is not "which model is smarter." It is "where does paying for Max stop earning its keep." Run the cheaper tier first and promote only the tasks that fail.

Choose Qwen-Plus when the task is general chat, summarization, structured extraction, or classification where a competent answer is the bar and a brilliant one adds nothing. At a routed $0.26 / $0.78, Plus handles these at roughly a third of the Max rate, and our live call confirms a real request lands near a ten-thousandth of a cent. Reserve Qwen3-Max for hard reasoning, long-context synthesis, or code where a wrong answer is expensive to catch downstream.

A practical rule we use: if a human reviewer cannot tell the Plus answer from the Max answer on a sample of your real prompts, the quality premium is imaginary and you are paying five times the output rate for nothing. Test on your own traffic before committing a tier, because the crossover depends on your prompts, not on a benchmark leaderboard.

How Qwen-Plus compares to DeepSeek

Buyers shortlisting a mid-tier Chinese model usually weigh Qwen-Plus against DeepSeek's general endpoint, so the cross-vendor number matters. According to DeepSeek API Docs, DeepSeek V4-Flash bills $0.14 input and $0.28 output per million tokens on its official endpoint, which undercuts the routed Qwen-Plus rate on both sides.

The comparison carries an asymmetry we have to flag: the DeepSeek figure is the official first-party rate, while the Qwen-Plus figure is OpenRouter-routed. We are not comparing two official endpoints; we are comparing the route each model is realistically reached on in this environment. On raw published rates DeepSeek is cheaper, but Qwen-Plus brings Alibaba's tooling, regional endpoints, and a different training mix. The full rate card and routing notes live in our Qwen API pricing hub, and the DeepSeek side is broken out in the DeepSeek API pricing hub.

If absolute lowest token price is the only criterion, DeepSeek V4-Flash wins this pair. If you want Alibaba's ecosystem or already route Qwen elsewhere in your stack, Plus is the tier to standardize on before Max.

FAQ

How much does Qwen-Plus cost per million tokens? Measured via OpenRouter on 2026-06-26, Qwen-Plus routes at $0.26 input and $0.78 output per million tokens. Alibaba's own published rate is higher, roughly $0.40 input and $1.20 output, per Alibaba Cloud Model Studio pricing, pending native re-verification on a DashScope key.

Is Qwen-Plus cheaper than Qwen3-Max? Yes, substantially. The routed Plus rate ($0.26 / $0.78) is about a third of the routed Max rate ($0.78 / $3.90). In our live test the same prompt cost $0.0000832 on Plus versus $0.00027534 on Max, over three times more, despite Max returning a shorter answer.

What did a real Qwen-Plus call cost in your test? One live call routed via OpenRouter (served as qwen/qwen-plus-2025-01-25) billed $0.0000832 for 38 input and 94 output tokens, returning in 2.47 seconds. Output tokens drove most of the cost, which is why the output rate matters more than input for generative work.

Should I use Qwen-Plus or DeepSeek? On published rates, DeepSeek V4-Flash ($0.14 / $0.28 official) is cheaper than routed Qwen-Plus. Choose DeepSeek for lowest token price; choose Qwen-Plus if you want Alibaba's ecosystem or already route Qwen. Note the comparison mixes a first-party DeepSeek rate with an OpenRouter-routed Qwen rate.

---

This is part of the Qwen API pricing hub, where the full cluster rate card and routing methodology live.

Author: Kevin Fan, Customer Success Manager at China LLM Directory, specializing in Chinese LLM ecosystem pricing. Last verified: 2026-06-26.

<!-- METADATA { "title": "Qwen-Plus API Pricing: The Workhorse Tier (2026)", "slug": "qwen-plus-pricing", "meta_description": "Qwen-Plus routes at $0.26/$0.78 per 1M via OpenRouter vs ~$0.40/$1.20 official. Live call: 38/94 tokens, $0.0000832, 2.47s. When Plus beats Max on cost. 2026.", "focus_keyword": "qwen-plus api pricing", "secondary_keywords": ["qwen-plus cost per million tokens", "qwen plus vs qwen max price", "qwen plus openrouter price", "qwen plus vs deepseek pricing"], "tags": ["Qwen", "API Pricing", "Alibaba Cloud"], "category": "Pricing", "cluster_id": "qwen-api-pricing", "cluster_role": "micro", "hub_slug": "qwen-api-pricing", "evidence_file": "clients/china-llm-aggregator/articles/qwen-api-pricing-evidence.json", "needs_native_reverify": true, "verified_until": "2026-09-24", "author_name": "Kevin Fan", "author_title": "Customer Success Manager", "author_linkedin": "", "author_expertise": ["Chinese LLM ecosystem", "AI infrastructure pricing", "model benchmarking", "cross-border AI compliance"], "faq_pairs": [ {"q": "How much does Qwen-Plus cost per million tokens?", "a": "Measured via OpenRouter on 2026-06-26, Qwen-Plus routes at $0.26 input and $0.78 output per million tokens. Alibaba's own published rate is higher, roughly $0.40 input and $1.20 output, per Alibaba Cloud Model Studio pricing, pending native re-verification on a DashScope key."}, {"q": "Is Qwen-Plus cheaper than Qwen3-Max?", "a": "Yes, substantially. The routed Plus rate ($0.26 / $0.78) is about a third of the routed Max rate ($0.78 / $3.90). In our live test the same prompt cost $0.0000832 on Plus versus $0.00027534 on Max, over three times more, despite Max returning a shorter answer."}, {"q": "What did a real Qwen-Plus call cost in your test?", "a": "One live call routed via OpenRouter (served as qwen/qwen-plus-2025-01-25) billed $0.0000832 for 38 input and 94 output tokens, returning in 2.47 seconds. Output tokens drove most of the cost, which is why the output rate matters more than input for generative work."}, {"q": "Should I use Qwen-Plus or DeepSeek?", "a": "On published rates, DeepSeek V4-Flash ($0.14 / $0.28 official) is cheaper than routed Qwen-Plus. Choose DeepSeek for lowest token price; choose Qwen-Plus if you want Alibaba's ecosystem or already route Qwen. Note the comparison mixes a first-party DeepSeek rate with an OpenRouter-routed Qwen rate."} ], "external_links_used": [ {"url": "https://www.alibabacloud.com/help/en/model-studio/model-pricing", "source_name": "Alibaba Cloud Model Studio pricing", "claim": "Qwen-Plus official rate ~$0.40 input / $1.20 output per 1M tokens (needs native re-verify)"}, {"url": "https://api-docs.deepseek.com/quick_start/pricing/", "source_name": "DeepSeek API Docs – Pricing", "claim": "DeepSeek V4-Flash official rate $0.14 input / $0.28 output per 1M tokens (comparison anchor)"} ], "internal_links_used": [ {"url": "/blog/qwen-api-pricing/", "anchor_text": "Qwen API pricing hub", "type": "hub"}, {"url": "/blog/deepseek-api-pricing/", "anchor_text": "DeepSeek API pricing hub", "type": "cross-cluster"} ], "first_hand_evidence": { "source": "qwen-api-pricing-evidence.json run label qwen-plus_general (+ qwen3-max_general for contrast)", "measured": "Qwen-Plus via OpenRouter, served qwen/qwen-plus-2025-01-25: 38 prompt + 94 completion tokens, billed $0.0000832, latency 2.47s. Contrast Qwen3-Max general: 38/63, $0.00027534, 3.22s.", "routing_disclosure": "OpenRouter-routed; not Alibaba official endpoint", "captured": "2026-06-26" }, "images_status": "spec-only (not generated; FAL_API_KEY unset)", "images": [ {"position": "featured", "type": "generated", "prompt": "Clean editorial diagram comparing three Qwen API price tiers per million tokens, with Qwen-Plus highlighted in the middle as the workhorse tier between cheap small models and the Max flagship. Annotate the live measured cost of $0.0000832 for a 38-in/94-out call routed via OpenRouter. Blue and amber palette, minimal background, 16:9.", "alt": "Diagram comparing Qwen API price tiers per million tokens with Qwen-Plus highlighted as the mid workhorse tier, annotated with the live measured cost of $0.0000832 for a 38-input 94-output call"} ] } -->