Skip to main content

Qwen3 Max

Qwen3-Max is an updated release built on the Qwen3 series, offering major improvements in reasoning, instruction following, multilingual support, and long-tail knowledge coverage compared to the January 2025 version. It delivers higher accuracy in math, coding, logic, and science tasks, follows complex instructions in Chinese and English more reliably, reduces hallucinations, and produces higher-quality responses for open-ended Q\u0026A, writing, and conversation. The model supports over 100 languages with stronger translation and commonsense reasoning, and is optimized for retrieval-augmented generation (RAG) and tool calling, though it does not include a dedicated “thinking” mode.

Input / output price
$1.20 / $6.00 per 1M
Context
256K tokens
Released
Sep 23, 2025
Agent setup

Paste into Claude Code, Cursor, or Codex

Help me set up the Respan gateway: fetch and follow https://www.respan.ai/docs/agent-setup.md. Use model "qwen3-max".

Read docs

Providers

One provider serves this model, priced per 1M tokens after gateway discounts.

ProviderInputOutput
Concentrateconcentrate/qwen3-max
$1.20/M$6.00/M

Pricing

USD per 1M tokens on Concentrate, after gateway discounts.

Input
$1.20 per 1M
Output
$6.00 per 1M
Cache read
$0.24 per 1M

Cost by volume

Total cost from 0 to 1T tokens.

$1.20/MConcentrate

Throughput and latency

30-day averages from requests through the Respan gateway. Under each figure, every model on the gateway by its average, this one in blue.

Output speed

Higher is better

13tokens/sacross providers

Faster than 39% of the other 595 models on the gateway

Response latency

Lower is better

2.6sacross providers

Faster than 41% of the other 756 models on the gateway

Benchmarks

The Respan Index plus 10 published scores from 4 sources, each linked to where it was published.

19.6/ 100

Respan Index#131 of 176

176 ranked models. The shaded band is this model's 95% interval, 16.3 to 26.5. Claude Haiku 4.5 and Claude 3.7 Sonnet fall inside it, so the gap is within the margin of error.
Qwen3 Max by category: Coding 4.8, Agents 5.1, Reasoning 20.1, Knowledge & language 48.3, Human preference 48.3
  1. 131Qwen3-Max19.6
  2. 132Claude Haiku 4.519.3
  3. 132Claude 3.7 Sonnet19.3

Published scores

BenchmarkScore
SWE-bench VerifiedQwen69.6%

Compare Qwen3 Max: Qwen3-Max vs Qwen3.7-Max, Qwen3-Max vs Qwen3.8-Max, Qwen3-Max vs Qwen3.8-Flash-Next, Qwen3-Max vs Qwen3.8 27B, Qwen3-Max vs Qwen3.6-Max-Preview, Qwen3-Max vs Qwen3.7-Plus

Apps

Use Qwen3 Max in the tools you already run. Each guide points the app at the Respan gateway.

Setup guides for more apps are in the integration docs.

Qwen3 Max FAQ

Short answers, built from the data on this page.

Qwen3 Max is a model from Qwen, released on Sep 23, 2025. Qwen3-Max is an updated release built on the Qwen3 series, offering major improvements in reasoning, instruction following, multilingual support, and long-tail knowledge coverage compared to the January 2025 version. It delivers higher accuracy in math, coding, logic, and science tasks, follows complex instructions in Chinese and English more reliably, reduces hallucinations, and produces higher-quality responses for open-ended Q\u0026A, writing, and conversation. The model supports over 100 languages with stronger translation and commonsense reasoning, and is optimized for retrieval-augmented generation (RAG) and tool calling, though it does not include a dedicated “thinking” mode.

On the Respan gateway, Qwen3 Max costs $1.20 per 1M input tokens and $6.00 per 1M output tokens on Concentrate, after gateway discounts.

Qwen3 Max is served by Concentrate.

Qwen3 Max accepts up to 256K tokens of context.

Point any OpenAI-compatible SDK at https://gateway.respan.ai/v1 with your Respan API key and set the model to "qwen3-max". To set it up with a coding agent, copy the agent setup prompt at the top of this page.

Qwen3 Max scores 19.6 out of 100 on the Respan Index v1.1, ranking #131 of 176 models. By category: Coding 4.8, Agents 5.1, Reasoning 20.1, Knowledge & language 48.3, and Human preference 48.3. The Respan Index combines independent and lab-reported benchmark results into one score, using benchmark data collected by Respan.

Qwen3 Max has 10 published benchmark scores, including SWE-bench Verified 69.6%, Vending-Bench 2 71.57, and GPQA Diamond 72.6%. Each score links to where it was published; the data is collected by Respan.

Certified where it counts.

Built to meet the security and privacy standards that enterprise and healthcare teams require.

ISO 27001

The internationally recognized standard for information security management.

SOC 2

Secure, compliant management of your data across all of our systems.

GDPR

Operated under GDPR, the world's strictest standard for data privacy.

HIPAA

HIPAA compliant, with a BAA available for healthcare teams.

Works with your entire stack

Use Respan with your favorite frameworks and tools.

AnthropicAI21 LabsxAIVercel AI SDK
LangChainOpenRouterAssemblyAICursor
LiteLLMGroqBasetenMem0
CohereCogneeLlamaIndexFireworks
DeepSeekLinkupHaystackTogether AI
PerplexityInference.netPostHogRubyLLM
SlackMastraAzure OpenAIMistral
AWS BedrockMoonshotHyperspellBAML
ModaInstructorGoogle GeminiParasail
ReductoZapierAgnoNebius
OpenAINovita AIReplicateSuperagent

Built for AI agents.
Break less.
Ship more.