qwen3-next-80b-think¶
Tier 2: 27.5 points · 1,000 tokens (500 input + 500 output)
Comparison method and assumptions.
Pricing¶
API Pricing (USD)¶
| Type | Cost |
|---|---|
| Prompt | $1.515E-7/token |
| Completion | $0.0000015152/token |
Website Pricing¶
| Type | Cost |
|---|---|
| Input Text | 5 points/1k tokens |
| Output (Text) | $1.52/1M tokens · 50 points/1k tokens |
All parsed rates (both currencies):
| Service | Currency | Amount | Per |
|---|---|---|---|
| Input (text) | USD | 0.15 | 1000000 tokens |
| Input (text) | POINTS | 5 | 1000 tokens |
| Output (text) | USD | 1.52 | 1000000 tokens |
| Output (text) | POINTS | 50 | 1000 tokens |
Last Checked: 2026-10-22 13:26:25.467350
Bot Information¶
Creator: @novitaai
Description: Try https://poe.com/Qwen3.8-Flash for the first-party Qwen bot.
Qwen3-Next uses a highly sparse MoE design: 80B total parameters, but only ~3B activated per inference step. Experiments show that, with global load balancing, increasing total expert parameters while keeping activated experts fixed steadily reduces training loss.Compared to Qwen3’s MoE (128 total experts, 8 routed), Qwen3-Next expands to 512 total experts, combining 10 routed experts + 1 shared expert — maximizing resource usage without hurting performance. The Qwen3-Next-80B-A3B-Thinking excels at complex reasoning tasks — outperforming higher-cost models like Qwen3-30B-A3B-Thinking-2507 and Qwen3-32B-Thinking, outpeforming the closed-source Gemini-2.5-Flash-Thinking on multiple benchmarks, and approaching the performance of our top-tier model Qwen3-235B-A22B-Thinking-2507.
File Support: Text, Markdown and PDF files Context window: 131k tokens
Extra: Powered by a server managed by @novitaai. Learn more
Architecture¶
Input Modalities: text
Output Modalities: text
Modality: text->text
Technical Details¶
Model ID: qwen3-next-80b-think
Object Type: model
Created: 1757556610505
Owned By: Novita AI
Root: qwen3-next-80b-think
API Last Updated: 2026-10-22 13:58:59.075129