by Alibaba/Qwen· 9 months ago
Alibaba's large-scale reasoning model with ~1T parameters and chain-of-thought capabilities.
Context Window
256K
Max Output
16K
TTFT
1000ms
Speed
50 tok/s
Input Price
$1.20/M tokens
Output Price
$6.00/M tokens
Performance Profile
Frontier-tier performance at $1.20/M input tokens
256K token context window — handles lengthy documents with ease
Supports text + code — true multimodal capability
~1T (22B active) parameter architecture for deep reasoning
vs similar-tier models
| Model | Input | Output | Context | Avg Score |
|---|---|---|---|---|
Qwen3 Max ThinkingCurrent Alibaba/Qwen | $1.20 | $6.00 | 256K | 72.7 |
GPT-4o OpenAI | $2.50 | $10.00 | 128K | 81.1 |
Kimi K2.5 Moonshot AI | $0.45 | $2.20 | 256K | 92.3 |
Generate a function
$0.0024Spec → implementation with tests
500 in · 300 out
Review a 2,000-line PR
$0.024Full pull request code review
10,000 in · 2,000 out
Refactor a 5,000-line module
$0.060Major refactoring with explanations
25,000 in · 5,000 out
Analyze a full codebase
$0.180Architecture analysis + recommendations
100,000 in · 10,000 out
Code generation
$72/mo
$2/day
PR reviews
$720/mo
$24/day
Codebase analysis
$3240/mo
$108/day
No ratings yet. Be the first to rate this model!
Sign in to rate this model and share your experience.
Sign in to leave a comment and join the discussion.
Alibaba/Qwen
Specialized code model trained on 7.5T tokens (70% code). Supports 100+ programming languages and agentic workflows.
Input
$0.30/M
Output
$0.60/M
Context
262K
Alibaba/Qwen
Most capable open VLM rivaling GPT-5 across multimodal benchmarks. Strong reasoning and agentic capabilities.
Input
$0.30/M
Output
$0.60/M
Context
128K
Alibaba/Qwen
Alibaba's open-weight hybrid MoE model with 512 experts and 17B active parameters. Natively multimodal with 201 language support. Top scores on GPQA and SWE-bench.
Input
$0.15/M
Output
$1.00/M
Context
256K
OpenAI
OpenAI's most advanced multimodal model. Excels at text, vision, and audio tasks with fast response times.
Input
$2.50/M
Output
$10.00/M
Context
128K
Moonshot AI
Moonshot AI's frontier multimodal MoE model with 1T total parameters (32B active). Tops SWE-bench and AIME 2025 benchmarks.
Input
$0.45/M
Output
$2.20/M
Context
256K
Google's most capable thinking model with breakthrough performance on reasoning and coding.
Input
$1.25/M
Output
$10.00/M
Context
1.0M